1 of 178

Seasonal influences on hydrologic anomalies �in headwater catchments

Daniel Hogan�PhD Dissertation Proposal

March 10, 2025�Civil and Environmental Engineering�University of Washington ��Supervisory Committee:

�Jessica D. Lundquist, Chair �Erkan Istanbulluoglu, Member � Civil and Environmental Engineering, UW �Bart Nijssen, Member � Civil and Environmental Engineering, UW �Lynn McMurdie, Graduate School Rep.� Atmospheric Sciences, UW

Rosemary Carroll, Member, � Desert Research Institute ��

NSF Award No. 2139836

2 of 178

Motivation – The importance of snow-dominated catchments

Drinking

Water

#2

Photo by

Jeremy Snyder

Agriculture

Hydropower

Mountain ecosystems

Snow-dominated catchments provide water for…

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

3 of 178

Motivation – Western US water supply relies on mountain snowpack

Snow provides upwards of 60% of water to streamflow (Li et al., 2017)

Snowpack-streamflow relationship has guided water forecasts for over a century (Sturm et al., 2015)

April 1 snow water equivalent (SWE) is a primary predictor for streamflow

#3

Figure 1(a) from Li et al. (2017)

James Church

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

4 of 178

Motivation – Western US water supply relies on mountain snowpack

Snow provides upwards of 60% of water to streamflow (Li et al., 2017)

Snowpack-streamflow relationship has guided water forecasts for over a century (Sturm et al., 2015)

April 1 snow water equivalent (SWE) is a primary predictor for streamflow

#4

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

5 of 178

Motivation – Western US water supply relies on mountain snowpack

Snow provides upwards of 60% of water to streamflow (Li et al., 2017)

Snowpack-streamflow relationship has guided water forecasts for over a century (Sturm et al., 2015)

April 1 snow water equivalent (SWE) is a primary predictor for streamflow

#5

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

6 of 178

Motivation – Consequences of anomalies to water supply forecasts

Recent decreases in water supply forecast performance led to hydrologic anomalies�(Pagano et al., 2005)

#6

SWE predicts streamflow, but what causes anomalies?

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

7 of 178

Motivation – Consequences of anomalies to water supply forecasts

Recent decreases in water supply forecast performance led to hydrologic anomalies�(Pagano et al., 2005)

#7

Figure 2 in Goble and Schumacher (2023) from 2021 NRCS forecast

Snowpack �(near normal)

Forecast �(near normal)

Observed �(far below normal)

Snowpack �(near normal)

Forecast �(near normal)

SWE predicts streamflow, but what causes anomalies?

Colorado

Utah

Colorado

Utah

Colorado

Utah

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

8 of 178

Motivation – Consequences to water supply forecasts

Hydrologic anomalies occur when observed streamflow falls far outside expectations

Recent examples in Colorado & California �(Goble & Schumacher, 2023; Lapides et al., 2022)
Several hypotheses posed �explain anomalies

#8

SWE predicts streamflow, but what causes anomalies?

Figure 2 in Goble and Schumacher (2023) from 2021 NRCS forecast

Snowpack �(near normal)

Forecast �(near normal)

Observed �(far below normal)

Snowpack �(near normal)

Forecast �(near normal)

SWE predicts streamflow, but what causes anomalies?

Colorado

Utah

Colorado

Utah

Colorado

Utah

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

start with a water resource manager for water district. They knows other processes are important, but with a normal snowpack he's confident water resources should be near normal for his district. April rolls around, snow obs are taken and forecasts integrate this information to project river flows for the summer.

Slide: but something is right… as summertime closes in, projections dwindle lower and lower until they sit at a fraction of where they were, just a few months before.
Slide: define as instances where, due to unaccounted or poorly represented processes, observed water resources fall far outside expectations.
Slide: Illustrate simple example with snow-streamflow relationship. Highlight anomalies
Slide: this does not happen every year, but in a warming climate, large year-over-year deficits have developed throughout the West

Define hydrologic anomalies. Walk through example of water manager

See's snowpack is normal. We have strong relationships, we highlight anomalies.

End with water manager and angry / worried users who don't know why things ended differently than expected

9 of 178

Motivation – Which processes might play a role?

Increases in snow sublimation �(Xiao et al., 2018; Gordon et al., 2022)
Precipitation shift from snow to rain�(Berghuijs et al., 2014)
Earlier loss of reflective snow → greater evapotranspiration�(Foster et al., 2016; Hogan & Lundquist, 2024; Meira Neto et al., 2020; Milly & Dunne, 2020)
Change in spring precipitation�(Goble & Schumacher, 2023; Hogan & Lundquist, 2024)
Change in subsurface contributions over time�(Carroll et al., 2024)
Lower antecedent soil moisture �(Lapides et al., 2022)

#9

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

10 of 178

Motivation – Why focus on the anomalies?

Pros

Highlight gaps in our understanding
Promote reflection on underlying causes.
Offer a chance to better represent the natural system

Cons:

There is no universal answer��

#10

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

TLDR: these hydrologic anomalies make us question what processes we represent well and which ones we don't represent well. They make us want to figure out what key processes aren't well captured within our current modeling framework.

Why are hydrologic anomalies important to understand?

They only occur when youre wrong by a lot. Opportunity to step back and think: what might be causing this to oour? My understanding of the process isn't what I thought it was. Emphasize areas where our understanding of the hydrologic system needs improvement
Opportunity to learn from the natural system.

Figure of relationship and then point to other processes.

Emphasize simple example but highlight uncertainties.

Benefits of studying anomalies.

Drawbacks/challenges with this approach. There often is not one simple answer!

11 of 178

Objectives – How do we explore these anomalies?

Look into components of the water balance

#11

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

12 of 178

Objectives – Our representation of the water balance

#12

USGS Natural Water Cycle

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

13 of 178

Objectives – Our representation of the water balance

#13

Streamflow

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

14 of 178

Objectives – Our representation of the water balance

#14

Precipitation

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

15 of 178

Objectives – Our representation of the water balance

#15

Snow sublimation

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

16 of 178

Objectives – Our representation of the water balance

#16

Evapotranspiration

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

17 of 178

Objectives – Our representation of the water balance

#17

Subsurface storage change

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

18 of 178

Objectives – Our representation of the water balance

#18

Measurement + Modeling Errors

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

19 of 178

Objectives - So what processes are most important

#19

Chapter 1: characterizing large sublimation events

Chapter 1

Chapter 2

Chapter 3

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

20 of 178

Objectives - So what processes are most important

#20

Chapter 2: evaluating precipitation uncertainties across seasons

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

21 of 178

Observations from the SOS, SAIL, and SPLASH Campaigns

#21

Miller et al., 2021

NSF Award No. 2139836

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

22 of 178

Objectives - So what processes are most important

#22

Chapter 3: Iterating over model process representation

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

23 of 178

Objectives - So what processes are most important

#23

Chapter 4: Focus on effect of fall and spring conditions�

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

24 of 178

Objectives - So what processes are most important

#24

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Observations

Model experimentation & application

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

25 of 178

Chapter 1: What’s sublimation, why do we care?

Directly removes water from snowpack
Sublimation increases are hypothesized to contribute to recent streamflow decline �(Gordon et al., 2022; Xiao et al., 2018)

#25

Snow → Water vapor

Snowflake sublimating

Video courtesy of Kelly Elder

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

26 of 178

Chapter 1: What’s sublimation, why do we care?

Directly removes water from snowpack
Contributes to uncertainty in hydrologic models�(Xia et al., 2017; Slater et al., 2001)
Sublimation increases are hypothesized to contribute to recent streamflow decline �(Gordon et al., 2022; Xiao et al., 2018)

#26

Snow → Water vapor

Video courtesy of Kelly Elder

Snowflake sublimating

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

27 of 178

Sublimation is hard to predict

Uncertainty in how much:

Sublimation estimates range from 1% to 90% of seasonal snowfall locally �(Lundquist et al., 2024; Svoma et al., 2016)
Different measurement methods yield different results�(Sexstone et al., 2016)

Uncertainty in what drives it:

Sublimation increases often attributed directly to temperature rise �(Scaff et al., 2024; Harpold et al., 2012; Harpold & Brooks, 2018)
But, observations show periodic events control year-to-year variability�(Bliss et al., 2011; Hood et al. 1999; Knowles et al., 2012, Reba et al, 2012)

#27

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

As I just mentioned, sublimation is highly uncertain. Results from numerous studies report that sublimation can range from 1-90% locally.

This uncertainty means that constraining variability may be quite difficult.

This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

28 of 178

Sublimation is hard to predict

#28

Dec 1

Cumulative Sublimation

Mar 31

Uncertainty in how much:

Sublimation estimates range from 1% to 90% of seasonal snowfall locally �(Lundquist et al., 2024; Svoma et al., 2016)
Different measurement methods yield different results�(Sexstone et al., 2016)

Uncertainty in what drives it:

Sublimation increases often attributed directly to temperature rise �(Scaff et al., 2024; Harpold et al., 2012; Harpold & Brooks, 2018)
But, observations show periodic events control year-to-year variability�(Bliss et al., 2011; Hood et al. 1999; Knowles et al., 2012, Reba et al, 2012)

Observation �Period

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

As I just mentioned, sublimation is highly uncertain. Results from numerous studies report that sublimation can range from 1-90% locally.

This uncertainty means that constraining variability may be quite difficult.

This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

29 of 178

So, what can we do?

#29

Uncertainty in how much:

Sublimation estimates range from 1% to 90% of seasonal snowfall locally �(Lundquist et al., 2024; Svoma et al., 2016)

Uncertainty in what drives it:

Sublimation increases often attributed directly to temperature rise �(Scaff et al., 2024; Harpold et al., 2012; Harpold & Brooks, 2018)
But, observations show periodic events control year-to-year variability�(Bliss et al., 2011; Hood et al. 1999; Knowles et al., 2012, Reba et al, 2012)

We can take high quality, continuous measurements

Eddy covariance system �(PC: Emilio Mateo)

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

As I just mentioned, sublimation is highly uncertain. Results from numerous studies report that sublimation can range from 1-90% locally.

This uncertainty means that constraining variability may be quite difficult.

This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

30 of 178

So, what can we do?

#30

Uncertainty in how much:

Sublimation estimates range from 1% to 90% of seasonal snowfall locally �(Lundquist et al., 2024; Svoma et al., 2016)

Uncertainty in what drives it:

Sublimation increases often attributed directly to temperature rise �(Scaff et al., 2024; Harpold et al., 2012; Harpold & Brooks, 2018)
But, observations show periodic events control year-to-year variability�(Bliss et al., 2011; Hood et al. 1999; Knowles et al., 2012, Reba et al, 2012)

Eddy covariance system �(PC: Emilio Mateo)

We can determine the importance of periodic events and characterize them

We can take high quality, continuous measurements

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

As I just mentioned, sublimation is highly uncertain. Results from numerous studies report that sublimation can range from 1-90% locally.

This uncertainty means that constraining variability may be quite difficult.

This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

31 of 178

How can we do this?

We have 2 goals:

Continuously monitor sublimation
Determine how much large, periodic sublimation events contribute to total

#31

NSF Award No. 2139836

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Difficult to model in complex terrain

Its difficult to model because its representation relies on assumptions that are commonly invalidated in complex terrain. I won't go over these reasons, but here but Eli likely has some thoughts on the subject. These can break down in certain conditions common to these mountain environments.

Rarely measured in complex terrain. Measurement is also tricky. There are a few different methods used to estimate it in the field, but only really one way to measure it.
This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Not considered in the current USGS water cycle.

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

32 of 178

How can we do this?

Introducing the Sublimation of Snow campaign

We had a unique opportunity to continuously measure sublimation

#32

NSF Award No. 2139836

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Difficult to model in complex terrain

Its difficult to model because its representation relies on assumptions that are commonly invalidated in complex terrain. I won't go over these reasons, but here but Eli likely has some thoughts on the subject. These can break down in certain conditions common to these mountain environments.

Rarely measured in complex terrain. Measurement is also tricky. There are a few different methods used to estimate it in the field, but only really one way to measure it.
This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Not considered in the current USGS water cycle.

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

33 of 178

Research Objectives: Look at importance of large sublimation events

1. How much do these large events contribute to winter sublimation?

2. What characterizes these events?

3. How might we identify these events at larger-scales?

#33

NSF Award No. 2139836

Addressed by using data from 2 winter seasons

Dept. of Energy

NOAA

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

34 of 178

Research Objectives: Look at importance of large sublimation events

1. How much do these large events contribute to winter sublimation?

2. What characterizes these events?

3. How might we identify these events at larger-scales?

#34

NSF Award No. 2139836

Addressed by using data from 2 winter seasons

S³

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

35 of 178

Observations from the SOS, SAIL, and SPLASH Campaigns

#35

Miller et al., 2021

NSF Award No. 2139836

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

36 of 178

Observations: Site-scale, valley-scale, synoptic-scale

#36

Avery Picnic

Gothic

= Doppler lidar

= Flux station

= T/RH

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

NSF Award No. 2139836

= Radiosonde

= SEB

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

37 of 178

Observations: Site-scale, valley-scale, synoptic-scale

#37

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

= Doppler lidar

= Flux station

= T/RH

= Radiosonde

= SEB

= Blowing snow flux

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

38 of 178

Observations at the site-scale begin to tell our story

#38

Wind Speed (m/s)

Sublimation Rate (mm/hr)

Temperature (C)

Blowing snow flux (g/m2/s)

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

But how do you plan on doing this?

We measured sublimation. It relates to certain variables. We see two different categories. We want to focus on the large events circeld.
I keep referring to these "large events", but I need to define what a "large sublimation event" really is.

We want to find periods where sublimation is elevated for a certain duration.
We define these events as a balance between intensity and duration. Our data is recorded inititally at a very high rate that is then averaged to 5-minute windows.
These 5-minute windows are too fine to get ahold of the bigger picture, so we resampled our data to a coarser temporal resolution to try and find periods where sublimation was consistently elevated.
We identified a 3-hour averaging period as a useful length to balance elevated intensity for a relatively long period of time.
We wanted to remove spikes and fluctuations that may balance out over a longer time period.
Doing this, we found periods clustered into two different groups. Short (<12 Hour), intense periods, where sublimation rates were greater than the 95th percentile of 3-hour observations and long duration events where values were elevated for >12 hours above the 75th percentile.

Identifying events. Show time series of 2023 observations.

Circle events of interest. What’s their effect, how do they stick out?

Build cartoon of intensity vs. duration. Describe how we try to capture events.

Show figure of intensity and duration plot built up sequentially as I describe it.

39 of 178

We identified two “regimes” where sublimation rates are elevated

#39

Wind Speed (m/s)

Sublimation Rate (mm/hr)

Temperature (C)

Blowing snow flux (g/m2/s)

Are these events continuous? Or sporadic?

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

But how do you plan on doing this?

We measured sublimation. It relates to certain variables. We see two different categories. We want to focus on the large events circeld.
I keep referring to these "large events", but I need to define what a "large sublimation event" really is.

We want to find periods where sublimation is elevated for a certain duration.
We define these events as a balance between intensity and duration. Our data is recorded inititally at a very high rate that is then averaged to 5-minute windows.
These 5-minute windows are too fine to get ahold of the bigger picture, so we resampled our data to a coarser temporal resolution to try and find periods where sublimation was consistently elevated.
We identified a 3-hour averaging period as a useful length to balance elevated intensity for a relatively long period of time.
We wanted to remove spikes and fluctuations that may balance out over a longer time period.
Doing this, we found periods clustered into two different groups. Short (<12 Hour), intense periods, where sublimation rates were greater than the 95th percentile of 3-hour observations and long duration events where values were elevated for >12 hours above the 75th percentile.

Identifying events. Show time series of 2023 observations.

Circle events of interest. What’s their effect, how do they stick out?

Build cartoon of intensity vs. duration. Describe how we try to capture events.

Show figure of intensity and duration plot built up sequentially as I describe it.

40 of 178

Periodic events control sublimation totals

#40

Two large event “types” balance duration & intensity

Long (>12 hours)
Short, intense (<12 hours)

Similar sublimation totals between years �(5-10% of peak SWE)
Large events dominate

>50% of winter sublimation occurs in just ~10% of time

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

But how do you plan on doing this?

We measured sublimation. It relates to certain variables. We see two different categories. We want to focus on the large events circeld.
I keep referring to these "large events", but I need to define what a "large sublimation event" really is.

We want to find periods where sublimation is elevated for a certain duration.
We define these events as a balance between intensity and duration. Our data is recorded inititally at a very high rate that is then averaged to 5-minute windows.
These 5-minute windows are too fine to get ahold of the bigger picture, so we resampled our data to a coarser temporal resolution to try and find periods where sublimation was consistently elevated.
We identified a 3-hour averaging period as a useful length to balance elevated intensity for a relatively long period of time.
We wanted to remove spikes and fluctuations that may balance out over a longer time period.
Doing this, we found periods clustered into two different groups. Short (<12 Hour), intense periods, where sublimation rates were greater than the 95th percentile of 3-hour observations and long duration events where values were elevated for >12 hours above the 75th percentile.

Identifying events. Show time series of 2023 observations.

Circle events of interest. What’s their effect, how do they stick out?

Build cartoon of intensity vs. duration. Describe how we try to capture events.

Show figure of intensity and duration plot built up sequentially as I describe it.

41 of 178

Defining large sublimation events

#41

We want to find periods where sublimation is elevated for a certain duration.
We define these events as a balance between intensity and duration. Our data is recorded inititally at a very high rate that is then averaged to 5-minute windows.
These 5-minute windows are too fine to get ahold of the bigger picture, so we resampled our data to a coarser temporal resolution to try and find periods where sublimation was consistently elevated.
We identified a 3-hour averaging period as a useful length to balance elevated intensity for a relatively long period of time.

Duration (hours)

Intensity (sublimation percentile)

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

But how do you plan on doing this?

We measured sublimation. It relates to certain variables. We see two different categories. We want to focus on the large events circeld.
I keep referring to these "large events", but I need to define what a "large sublimation event" really is.

We want to find periods where sublimation is elevated for a certain duration.
We define these events as a balance between intensity and duration. Our data is recorded inititally at a very high rate that is then averaged to 5-minute windows.
These 5-minute windows are too fine to get ahold of the bigger picture, so we resampled our data to a coarser temporal resolution to try and find periods where sublimation was consistently elevated.
We identified a 3-hour averaging period as a useful length to balance elevated intensity for a relatively long period of time.
We wanted to remove spikes and fluctuations that may balance out over a longer time period.
Doing this, we found periods clustered into two different groups. Short (<12 Hour), intense periods, where sublimation rates were greater than the 95th percentile of 3-hour observations and long duration events where values were elevated for >12 hours above the 75th percentile.

Identifying events. Show time series of 2023 observations.

Circle events of interest. What’s their effect, how do they stick out?

Build cartoon of intensity vs. duration. Describe how we try to capture events.

Show figure of intensity and duration plot built up sequentially as I describe it.

42 of 178

Large events combine to account for >50% of winter sublimation

#42

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

43 of 178

Large events combine to account for >50% of winter sublimation

#43

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

44 of 178

Large events combine to account for >50% of winter sublimation

#44

Two event characteristics

Long events occur around precipitation�
Short events occur later in season

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

45 of 178

Large events combine to account for >50% of winter sublimation

#45

Two event characteristics

Long events occur around precipitation�
Short events occur later in season

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

46 of 178

Results - Site-scale, valley-scale, synoptic-scale

#46

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

47 of 178

Results - Site-scale, valley-scale, synoptic-scale

#47

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Energy sources differ between event types

Flux stations

Surface energy balance

��Diurnal cycle (solar radiation) drives short, intense events

��Strong winds & blowing snow drive long events

Blowing snow sensor

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

48 of 178

Results - Site-scale, valley-scale, synoptic-scale

#48

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

49 of 178

Results - Site-scale, valley-scale, synoptic-scale

#49

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Long duration events:

��Upper air & surface conditions well connected

�

Diurnal cycle dominates surface conditions

Valley-scale turbulence varies by event type

COLD

WARM

Valley inversion strength and depth differ by event type

Long duration events:

Short, intense events:

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

50 of 178

Observations: Site-scale, valley-scale, synoptic-scale

#50

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Long duration events:

Weak valley inversion in AM & PM�+

Strong turbulent signal within the valley

=

Upper air & surface conditions well connected��

Short, intense events:

Strong inversion in AM, weak in PM�+

Weaker turbulent signal within the valley

=

Diurnal cycle dominates surface conditions��

COLD

WARM

COLD

WARM

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

51 of 178

Results - Site-scale, valley-scale, synoptic-scale

#51

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

52 of 178

Results - Site-scale, valley-scale, synoptic-scale

#52

Valley-scale

Synoptic-scale

Site-scale

~1000 m

~10,000 m

~10 m

~1 m

Long duration events

�Conditions aloft include:

High winds�
Dry air

Often after precipitation.

�Events with these characteristics occur between 1-12x per winter (between 1979-2023)

%

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

53 of 178

Synoptic-scale observations: Radiosondes and Reanalysis

#53

Radiosondes ID dry, windy anomalies at 500 mb are relate best to surface events��See how those change in time, both radiosondes and reanalysis

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

54 of 178

Chapter 1: summary

We constrained sublimation to 5-10% of peak SWE
Large sublimation events account for >50% of sublimation

Long events connect closely with conditions aloft
Short, intense events occur diurnally during spring-like conditions

Without capturing these events, estimates may be off

Now, we have a better idea which large-scale conditions affect surface sublimation.

#54

Preparing for submission to Journal of Hydrometeorology in Spring 2025

Chapter 1

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

The controls on winter sublimation largely come from major events that take up a small part of winter. These are outsized contributions.

Evidence supports large-scale conditions promote sublimation at the site-scale, linking these processes
Two main conditions, stormy windy, sunny in spring.
Large-scale signals can be used to identify these events.
Generally small contribution relative to winter time precip.

Changes in the frequency of these events may modulate winter sublimation, although we did not see that in our data (same amount in winter)
Changes in winter sublimation are not simply controlled by temperature rise. If these events are missed, sublimation totals may be out of whack.
Limitations: valley scale observations, we know that processes will be different in canopy covered terrain and higher terrain. But if these events impact valley, they likely have larger imapcts in other areas.
Preparing for submission early this spring

55 of 178

Objectives - So what processes are most important

#55

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

56 of 178

Chapter 2 - Seasonal precipitation measurement comparison

Why focus on measured precipitation?�

Its fundamental to water balance�
Uncertainty in it plays a big role in uncertainty of hydrologic models

#56

Snow crystals from Gothic, CO

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

57 of 178

Background – Many have worked to address these issues

In the past, there have been lots of comparisons of gridded datasets, which show pretty large differences
These are often compared to surface observations, which carry their own uncertainty
At the site-scale, it’s a bit more difficult and costs lots of time and money, so others that have done this have been limited in temporal length.
Conclusion from this is that we can get different estimates measuring precipitation can give different answers when measured in the same place.
Often focus on one season of precip, or events.
Still uncertainty when comparing different methods of measuring precipitation in different seasons.

#57

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What have people to address this?

In the past, there have been lots of comparisons of gridded datasets, which show pretty large differences
These are often compared to surface observations, which carry their own uncertainty
At the site-scale, it’s a bit more difficult and costs lots of time and money, so others that have done this have been limited in temporal length.
Conclusion from this is that we can get different estimates measuring precipitation can give different answers when measured in the same place.
Often focus on one season of precip, or events.
Still uncertainty when comparing different methods of measuring preciptiation in different seasons.

List comparison references and include some photos.

Highlight that these have been done in different ways.

Cover different windows

Pop up questions.

58 of 178

Motivation: similar to sublimation, there is still lots of uncertainty

Different ways of measuring precipitation give different results �(Rasmussen et al., 2011)

S³ field campaigns provide unique opportunity to compare methods over multiple seasons!

#58

I’ll get to explaining this plot soon

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

59 of 178

Research objectives:

We can take advantage of a unique dataset to:

Compare observations in time to assess biases.
Characterize events that caused big differences across seasons.

Other goals

We can compare to gridded radar product
We can compare to commonly used gridded datasets

#59

“Santa Slammer” �Atmospheric River Event

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are we doing here? Research objectives:

Well, we can take advantage of a unique dataset, where a bunch of gauges were set out in a mountain valley (East River). We'll look at 2 aspects
We can compare these observations in time to try to understand what processes lead to differences. How are they different at different times of year?
We can look at the characteristics of events that caused big differences across seasons.
We can scale up observations by using radar.
We can compare commonly used gridded datasets to their uncertainty relative to these observations.

Introduce observations from SAIL, SPLASH, SOS.

Start with Basin. Close in on UERB. Show icons for all the measurements and their location

Add list of different storms types that we can evaluate differences

Big inland penetrating AR
Blowing snow event
Summertime convection
Mixed precip.

60 of 178

Research objectives:

We can take advantage of a unique dataset to:

Compare observations in time to assess biases.
Characterize events that caused big differences across seasons.

Other goals

We can compare to gridded radar product
We can compare to commonly used gridded datasets

#60

But first, we need a benchmark

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are we doing here? Research objectives:

Well, we can take advantage of a unique dataset, where a bunch of gauges were set out in a mountain valley (East River). We'll look at 2 aspects
We can compare these observations in time to try to understand what processes lead to differences. How are they different at different times of year?
We can look at the characteristics of events that caused big differences across seasons.
We can scale up observations by using radar.
We can compare commonly used gridded datasets to their uncertainty relative to these observations.

Introduce observations from SAIL, SPLASH, SOS.

Start with Basin. Close in on UERB. Show icons for all the measurements and their location

Add list of different storms types that we can evaluate differences

Big inland penetrating AR
Blowing snow event
Summertime convection
Mixed precip.

61 of 178

Research objectives: benchmark observations

Why is his data our benchmark?

Same observation process, location, and observer since 1973
Accumulation gauge (rain + snow) validated against snow board
Twice daily observation & maintenance

#61

Introducing billy barr

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are we doing here? Research objectives:

Well, we can take advantage of a unique dataset, where a bunch of gauges were set out in a mountain valley (East River). We'll look at 2 aspects
We can compare these observations in time to try to understand what processes lead to differences. How are they different at different times of year?
We can look at the characteristics of events that caused big differences across seasons.
We can scale up observations by using radar.
We can compare commonly used gridded datasets to their uncertainty relative to these observations.

Introduce observations from SAIL, SPLASH, SOS.

Start with Basin. Close in on UERB. Show icons for all the measurements and their location

Add list of different storms types that we can evaluate differences

Big inland penetrating AR
Blowing snow event
Summertime convection
Mixed precip.

62 of 178

Observations from the SOS, SAIL, and SPLASH Campaigns

#62

Miller et al., 2021

NSF Award No. 2139836

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

63 of 178

Precipitation observations

#63

billy barr cabin

Gothic

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

Mt. Crested Butte (4km)

Potential Issues/limitations:

snow bridging,
gauge undercatch from wind,
no phase detection,
evaporative losses

Weighing/�storage gauge

2 km

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

64 of 178

Precipitation observations

#64

billy barr cabin

Gothic

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

Mt. Crested Butte (4km)

Potential Issues:

snow bridging,
gauge undercatch from wind,
tipping delays,
minimum detection limit

Tipping bucket

Weighing/�storage gauge

2 km

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

65 of 178

Precipitation observations

#65

billy barr cabin

Gothic

Snow pillow

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

Mt. Crested Butte (4km)

Potential Issues/Limitations:

Only useful when snow is present
Fully exposed to snow transport
Difficulty during melt season

Tipping bucket

Weighing/�storage gauge

2 km

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

66 of 178

Precipitation observations

#66

billy barr cabin

Gothic

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

Mt. Crested Butte (4km)

Potential Issues/Limitations:

Sensitive to blowing snow
Difficulty with frozen precipitation

Laser disdrometer

Tipping bucket

Weighing/�storage gauge

2 km

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

67 of 178

Precipitation observations

#67

billy barr cabin

Gothic

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

Mt. Crested Butte (4km)

Potential Issues/Limitations:

Not suited for frozen precip
Sensitive to blowing snow
Sensitive to high humidity

Laser disdrometer

Optical sensor

Tipping bucket

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

68 of 178

Precipitation observations

#68

billy barr cabin

Gothic

Laser disdrometer

Optical sensor

East River Valley

Gothic Mtn

Kettle Ponds

Courtesy of Eli Schwat

Tipping bucket

X-band radar

Mt. Crested Butte (4km)

Weighing/�storage gauge

Potential Issues/Limitations:

Sensitive to precip type – radar reflectivity relationship
Beam blockage

2 km

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

69 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#69

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

70 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#70

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

71 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#71

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

72 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#72

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

73 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#73

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

74 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#74

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

75 of 178

Results - Comparing winter precipitation shows large discrepancies

Variability on the order of 100% across instruments

#75

1 Dec 2021 – 31 Mar 2022

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

76 of 178

Results - Comparing winter precipitation shows large discrepancies

Winter 2022-2023 shows similar differences�

#76

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

1 Dec 2022 – 31 Mar 2023

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

77 of 178

Results - Is the mean error the problem?

Do these instruments have a simple mean error?

Or are there systematic errors?

#77

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Observation

Minus

Benchmark

Let’s assume no systematic errors: �use mean bias from winter 2021-2022 to predict winter 2022-2023

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

78 of 178

Results - Is the mean error the problem?

Accounting for mean bias helps,

but range is still ~30% �

#78

1 Dec 2022 – 31 Mar 2023 (mean bias corrected)

Laser disdrometer

Optical sensor

Tipping bucket

X-band radar

Weighing/�storage gauge

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

79 of 178

Results - Direct comparison shows large discrepancies

#79

Benchmark Gauge

Laser Disdrometer

LPDF Pluviometer

Kettle Ponds

SPLASH

Snow Pillow 1

Kettle Ponds

SOS

Snow Pillow 2

Kettle Ponds

SOS

Snow Pillow 3

Kettle Ponds

SOS

Snow Pillow 4

Kettle Ponds

SOS

Laser Disdrometer

Present Wx Detector

Radar

Optical Rain Gauge

Tipping Bucket

Weighing Bucket

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

80 of 178

Results - Direct comparison shows large discrepancies

#80

Benchmark Gauge

Laser Disdrometer

LPDF Pluviometer

Kettle Ponds

SPLASH

Snow Pillow 1

Kettle Ponds

SOS

Snow Pillow 2

Kettle Ponds

SOS

Snow Pillow 3

Kettle Ponds

SOS

Snow Pillow 4

Kettle Ponds

SOS

What causes kink?

Dec blowing snow event

What causes kink?

Early March divergence

Laser Disdrometer

Present Wx Detector

Radar

Optical Rain Gauge

Tipping Bucket

Weighing Bucket

What causes kinks?

Dec blowing snow event

What causes kinks?

Early season malfunction

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

81 of 178

Next Steps - evaluate systematic biases

#81

/

%

Compare precipitation over each season

Explore how different storm types influence biases

Evaluate gridded products against point & radar observations

Applying for funding through DOE SCGSR Program

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

82 of 178

Next Steps - evaluate systematic biases

#82

/

%

Compare precipitation over each season

Explore how different storm types influence biases

Applying for funding through DOE SCGSR Program

Evaluate gridded products against point & radar observations

Research ongoing, preparing for submission in Fall 2025

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

83 of 178

Objectives - So what processes are most important

#83

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Observations

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

84 of 178

Objectives - So what processes are most important

#84

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Model experimentation & application

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

85 of 178

Motivation – Some years, we are not forecasting streamflow well

Forecasts drive key water management decisions

#85

From: https://www.cbrfc.noaa.gov/wsup/graph/espgraph_hc.html?year=2025&id=ALEC2

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

86 of 178

Motivation – Some years, we are not forecasting streamflow well

Forecasts drive key water management decisions

#86

2021 CBRFC Water Supply Forecast for East River

Forecasted Aug 1

runoff (median)

Observed Aug 1 Runoff

Uncertainty

Average Aug 1 runoff

From: https://www.cbrfc.noaa.gov/wsup/graph/espgraph_hc.html?year=2025&id=ALEC2

Volume of Aug 1 Runoff

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

87 of 178

Motivation – Some years, we are not forecasting streamflow well

Forecasts drive key water management decisions

What drives these hydrologic anomalies?

#87

2021 CBRFC Water Supply Forecast for East River

Forecasted Aug 1

runoff (median)

Observed Aug 1 Runoff

Uncertainty

Average Aug 1 runoff

From: https://www.cbrfc.noaa.gov/wsup/graph/espgraph_hc.html?year=2025&id=ALEC2

SWE predicts streamflow, but what causes anomalies?

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

88 of 178

Motivation – Many hypotheses. How do we assess what matters?

Research has homed in on the importance of spring and fall conditions

Earlier loss of reflective snow → �greater ET
Change in spring precipitation
Lower antecedent soil moisture �

#88

SWE

We want to address these hypotheses, �but how?

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

89 of 178

Motivation – Many hypotheses. How do we assess what matters?

To test these hypotheses, we need a model.

Traditionally, a model is chosen…

�and parameters are tweaked to match a benchmark (like streamflow)…

But, research has shown that model structure is more important �(Henn et al., 2015)�

#89

From Clark et al., 2015

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

90 of 178

#90

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

91 of 178

Motivation – Finding a balance for process representation

#91

Less Complex�

More Complex�

SWE-Q regression

Distributed, �physically-based �model

From Ivanov et al. (2004)

Hydrologic Models

Bennett et al, 2019; �Vano et al., 2012

Missing important �processes

Difficult to decipher

We need to address hydrologic anomalies

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

92 of 178

Motivation – Finding a balance for process representation

#92

Less Complex�

More Complex�

Distributed, �physically-based �model

From Ivanov et al. (2004)

Hydrologic Models

Bennett et al, 2019; �Vano et al., 2012

Missing important �processes

Difficult to decipher

SWE-Q regression

What do we think is important� to represent?

Topographic

features

Snow Heterogeneity

Subsurface connectivity

We need to address hydrologic anomalies

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

93 of 178

Motivation – Finding a balance for process representation

#93

Structures for Unifying Multiple Modeling Alternatives (SUMMA)

One modeling system, many possibilities.
Think of it like Legos—building blocks (hydrologic response units or HRUs) let us test different hydrologic setups.
By iterating, we can refine what processes matter most.

Less Complex�

More Complex�

Missing important �processes

Difficult to decipher

Evaluate

Adjust

Develop

Made by DALL-E

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

94 of 178

Modeling locations, checkpoints and benchmarks

Ground our model using “checkpoints”

T/RH
Precip
ASO snow depth
Groundwater
Soil moisture

#94

Evaluate against streamflow benchmark

East River

Tuolumne River

East River

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Objectives

Well, we're going to use a model to try and build out these different processes in an iterative way, by adding in process complexity piece by piece to see which process representations make the biggest difference against our benchmarks,
while making sure we have our processes grounded in reality using observational networks in heavily instrumented basins.
They have these observations…
We're goin to focus on these three aspects in two different :

Representation of spatial discretization

Hypothesis 1

Representation of subsurface flow

Hypothesis 2

Representation of snow heterogeneity.

Hypothesis 3

How can we represent these differnences.

Center on 2 basins
Well isntrumented (summarize instrumentation with icons)
Focus on these aspects

1,2,3 list. With hypotheses. Add icon representation to illustrate

95 of 178

Motivation – Finding a balance for process representation

#95

Less Complex�

More Complex�

Missing important �processes

Difficult to decipher

Determine best forcing

Model forcing controls much uncertainty �(Tang et al. 2023; Raleigh et al. 2015)

Evaluate

Adjust

Develop

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

96 of 178

Motivation – Finding a balance for process representation

#96

Less Complex�

More Complex�

Missing important �processes

Difficult to decipher

Evaluate

Adjust

Develop

Incorporate different topographic features

Elevation dependence

Aspect dependence

Topographic factors

NEED TO DO ADD REFS

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

97 of 178

Motivation – Finding a balance for process representation

#97

Less Complex�

More Complex�

Missing important �processes

Difficult to decipher

Represent snow heterogeneity

Heterogeneity impacts soil moisture, ET, snow melt timing and runoff volume�(Newman et al., 2014, Badger et al., 2021)�

Aspect/Elevation

Patchy Snow

Evaluate

Adjust

Develop

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

“If we can represent additional key processes without overcomplicating things, we can better communicate results to decision-makers.”

(Badger et al., 2021; Luce et al., 1998; Lundquist & Dettinger, 2005; Newman et al., 2014),

Lundquist, J. D., & Dettinger, M. D. (2005). How snowpack heterogeneity affects diurnal
streamflow timing. Water Resources Research, 41(5).

Luce, C. H., Tarboton, D. G., & Cooley, K. R. (1998). The influence of the spatial distribution of

snow on basin-averaged snowmelt. Hydrological Processes, 12(10-11), 1671-1683.

Newman, A. J., Clark, M. P., Winstral, A., Marks, D., & Seyfried, M. (2014). The Use of

Similarity Concepts to Represent Subgrid Variability in Land Surface Models: Case

Study in a Snowmelt-Dominated Watershed. Journal of Hydrometeorology, 15(5), 1717-

Badger, A. M., Bjarke, N., Molotch, N. P., & Livneh, B. (2021). The sensitivity of runoff

generation to spatial snowpack uniformity in an alpine watershed: Green Lakes Valley,

Niwot Ridge Long-Term Ecological Research station. Hydrological Processes, 35(9),

98 of 178

Motivation – Finding a balance for process representation

#98

Less Complex�

More Complex�

Missing important �processes

Difficult to decipher

Test subsurface connectivity

Groundwater, once ignored, is now considered important �(Carroll et al., 2024)
High elevations export water to lower elevations�(Carroll et al., 2019; Fan et al., 2019; Kampf et al., 2020)
Contributes to ET variation�(Lundquist & Loheide II, 2011; Henn et al., 2018)

Evaluate

Adjust

Develop

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

99 of 178

Motivation – Finding a balance for process representation

#99

Less Complex�

More Complex�

Missing important �processes

Difficult to decipher

Evaluate

Adjust

Develop

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

100 of 178

Motivation – How do models represent these processes?

Models represent the hydrologic system in different ways. In physically based/distributed models, the fundamental unit is the hydrologic response unit. This can be a grid or some physical attribute of the system, but this the most basic model element.
We then connect the outputs from each of these HRUs. This is the plumbing of the system.
With this in mind, we might think through a few ways to represent this system.

Perhaps the only thing that matters is elevation, so we represent the system like that
Perhaps its also aspect dependent
Maybe it depends on the vegetation type in different locations
Or the snow depth at the beginning of the melt season
Or how subsurface moisture is routed through the system.

This offers a lot of questions. We need a setup that can represent these processes and let us evaluate what is most important

#100

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Let's walk through an example

Models represent the hydrologic system in different ways. In physically based/distributed models, the fundamental unit is the hydrologic response unit. This can be a grid or some physical attribute of the system, but this the most basic model element.
We then connect the outputs from each of these HRUs. This is the plumbing of the system.
With this in mind, we might think through a few ways to represent this system.

Perhaps the only thing that matters is elevation, so we represent the system like that
Perhaps its also aspect dependent
Maybe it depends on the vegetation type in different locations
Or the snow depth at the beginning of the melt season
Or how subsurface moisture is routed through the system.

This offers a lot of questions. We need a setup that can represent these processes and let us evaluate what is most important

Show a simple basin. Define HRU

Show grid over the basin, show another attribute like elevation

Then show how these are connected. The model does this. Plumbing analogy

Return to basin. Point down to different representation

Elevation matters?

Aspect matters?

Land cover type?

Snow depth in spring?

Subsurface moisture routing?

Lots of questions.

101 of 178

Results so far – Representation of elevation bands

#101

HRU 3

Streamflow

̄P = 350 mm

_

̄P = 275 mm

_

̄P = 200 mm

_

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

Forcing Data Applied to HRUs

Build Model Representation

Develop

Evaluate

Adjust

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

102 of 178

Results so far – Representation of elevation bands

#102

HRU 3

Streamflow

̄P = 350 mm

_

̄P = 275 mm

_

̄P = 200 mm

_

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

Forcing Data Applied to HRUs

Build Model Representation

Develop

Evaluate

Adjust

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

103 of 178

Next steps - Model iteration

#103

Forcing Data Applied to HRUs

Build Model Representation

HRU n

Streamflow

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

…

Groundwater

Soil Moisture

Develop

Evaluate

Adjust

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

104 of 178

Next steps - Model iteration

#104

Forcing Data Applied to HRUs

Build Model Representation

Develop

Evaluate

Adjust

HRU 3

Streamflow

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

HRU 6

Elevation Band

HRU 5

HRU 4

Groundwater

Soil Moisture

North Facing

South Facing

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

105 of 178

Next steps: Model iteration

#105

Forcing Data Applied to HRUs

Build Model Representation

Develop

Evaluate

Adjust

Streamflow

Elevation Band

Groundwater

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

106 of 178

Next steps – Model experimentation through iteration

Forcing selection
Iteratively add complexity

Aspect & Elevation: determine impacts from including topographic features
Snow Heterogeneity: develop statistical representation in SUMMA.
Subsurface configurations: test the effect of different subsurface configurations

#106

HRU 3

Streamflow

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

HRU 6

Elevation Band

HRU 5

HRU 4

Groundwater

Soil Moisture

North Facing

South Facing

Streamflow

Elevation Band

Groundwater

Soil Moisture

Working with team at NCAR with experimentation. Seeking to complete Spring 2026.

HRU n

Streamflow

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

…

Groundwater

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Whats next?

Forcing is important. I wont get into details here but we will be evaluating a few different alternatives to choose the best forcing.
We have the option to use many parameter to change representations. We can play around with parameter choice, but we want minimal tuning to try and see what we get and what we don’t get.
Work through this model setup, adding complexity iteratively.

Next do aspect + elevation
Build representation of snow cover heterogeneity in SUMMA (statistical representation)(Show examples of how we might do this)

Illustrate forcing required for SUMMA -> need to evaluate best inputs

Many parametsrs to choose from. Not the focus, but need to choose appropriate parameters for the processes we mention. How to illustrate this simply? Maybe just

Work through increasingly complex representations.

Show SUMMA figure to illustrate this ability. Turn on and off plumbing.

Show model representation from before with additional complexity.

Show snow cover heterogeneity ideas and walk through this

107 of 178

Chapter 4 – Spring and fall influences on streamflow

#107

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

108 of 178

Objectives – So what processes are most important

#108

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Model development & application

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

109 of 178

Motivation – Many hypotheses. How do we assess what matters?

Research has homed in on the importance of spring and fall conditions

Earlier loss of reflective snow → �greater ET
Change in spring precipitation
Lower antecedent soil moisture �

#109

SWE

We want to address these hypotheses, �but how?

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

110 of 178

Research objectives

Determine impacts of extreme spring and fall conditions (warm/cool, wet/dry) on runoff response
Isolate responses of different hydrologic mechanisms

ET,
soil moisture,
groundwater contributions,
snowmelt timing

#110

Made by DALL-E

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

111 of 178

Research plan

Testing hydrologic response with synthetic seasonal combinations

Create synthetic forcing by recombining extreme fall/spring conditions

Warm + dry, cool + wet

Assess how different combinations impact runoff efficiency and system states

#111

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

112 of 178

Research plan

Testing hydrologic response with synthetic seasonal combinations

Create synthetic forcing by recombining extreme fall/spring conditions

Warm + dry, cool + wet

Assess how different fall/spring combinations impact runoff efficiency and system states

#112

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

113 of 178

Research plan

Testing hydrologic response with synthetic seasonal combinations

Create synthetic forcing by recombining extreme fall/spring conditions

Extreme periods (e.g. May 2015)

Assess how different fall/spring combinations impact runoff efficiency and system states

#113

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

114 of 178

Different hypotheses – Fall Conditions

#114

Groundwater

Soil Moisture

Streamflow

Warm + Dry Fall

Cool + Wet Fall

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

115 of 178

Different hypotheses – Fall Conditions

#115

Groundwater

Soil Moisture

Streamflow

Warm + Dry Fall

Cool + Wet Fall

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

116 of 178

Different hypotheses – Fall Conditions

#116

Groundwater

Soil Moisture

Streamflow

Warm + Dry Fall

Cool + Wet Fall

Soil Moisture

ET

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

117 of 178

Different hypotheses – Spring Conditions

#117

Groundwater

Soil Moisture

Streamflow

Warm + Dry Spring

Cool + Wet Spring

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

118 of 178

Different hypotheses – Spring Conditions

#118

Groundwater

Soil Moisture

Streamflow

Warm + Dry Spring

Cool + Wet Spring

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

119 of 178

Different hypotheses – Spring Conditions

#119

Groundwater

Soil Moisture

Streamflow

Warm + Dry Spring

Cool + Wet Spring

Soil Moisture

ET

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

120 of 178

Current work and next steps

Analyze spring and fall anomalies, investigate events of interest
Construct synthetic forcing
Run model & evaluate hydrologic responses (soil moisture, GW, runoff, streamflow, ET)

#120

Working with team at NCAR with development. Seeking to complete Spring 2026.

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

121 of 178

Summary

#121

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Observations

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

122 of 178

Summary

#122

Chapter 1

Chapter 2

Chapter 4

Chapter 3

Model experimentation & application

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

123 of 178

Timeline

#123

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

124 of 178

124

Thank you for all those that have supported me along the way!

125 of 178

References

Abatzoglou, J. T., & Williams, A. P. (2016). Impact of anthropogenic climate change on wildfire across western US forests. Proceedings of the National Academy of Sciences, 113(42), 11770–11775. https://doi.org/10.1073/pnas.1607171113

Alam, S., Kaenel, M. von, Su, L., & Lettenmaier, D. P. (2024). The role of antecedent winter soil moisture carryover on spring runoff predictability in snow-influenced Western U.S. catchments. https://doi.org/10.1175/JHM-D-24-0038.1

Badger, A. M., Bjarke, N., Molotch, N. P., & Livneh, B. (2021). The sensitivity of runoff generation to spatial snowpack uniformity in an alpine watershed: Green Lakes Valley, Niwot Ridge Long-Term Ecological Research station. Hydrological Processes, 35(9), e14331. https://doi.org/10.1002/hyp.14331

Bales, R. C., Molotch, N. P., Painter, T. H., Dettinger, M. D., Rice, R., & Dozier, J. (2006). Mountain hydrology of the western United States. Water Resources Research, 42(8). https://doi.org/10.1029/2005WR004387

Bennett, A., Nijssen, B., Ou, G., Clark, M., & Nearing, G. (2019). Quantifying Process Connectivity With Transfer Entropy in Hydrologic Models. Water Resources Research, 55(6), 4613–4629. https://doi.org/10.1029/2018WR024555

Bliss, A. K., Cuffey, K. M., & Kavanaugh, J. L. (2011). Sublimation and surface energy budget of Taylor Glacier, Antarctica. Journal of Glaciology, 57(204), 684–696. https://doi.org/10.3189/002214311797409767

Carroll, R. W. H., Niswonger, R. G., Ulrich, C., Varadharajan, C., Siirila-Woodburn, E. R., & Williams, K. H. (2024). Declining groundwater storage expected to amplify mountain streamflow reductions in a warmer world. Nature Water, 2(5), 419–433. https://doi.org/10.1038/s44221-024-00239-0

Cline, D. W. (1997). Snow surface energy exchanges and snowmelt at a continental, midlatitude Alpine site. Water Resources Research, 33(4), 689–701. https://doi.org/10.1029/97WR00026

Clow, D. W. (2010). Changes in the Timing of Snowmelt and Streamflow in Colorado: A Response to Recent Warming. Journal of Climate, 23(9), 2293–2306. https://doi.org/10.1175/2009JCLI2951.1

Fan, Y. (2019). Are catchments leaky? WIREs Water, 6(6), e1386. https://doi.org/10.1002/wat2.1386

Goble, P. E., & Schumacher, R. S. (2023). On the Sources of Water Supply Forecast Error in Western Colorado. https://doi.org/10.1175/JHM-D-23-0004.1

Harpold, A., Brooks, P., Rajagopal, S., Heidbuchel, I., Jardine, A., & Stielstra, C. (2012). Changes in snowpack accumulation and ablation in the intermountain west. Water Resources Research, 48(11). https://doi.org/10.1029/2012WR011949

Harpold, A. A., & Brooks, P. D. (2018). Humidity determines snowpack ablation under a warming climate. Proceedings of the National Academy of Sciences, 115(6), 1215–1220. https://doi.org/10.1073/pnas.1716789115

Henn, B., Clark, M. P., Kavetski, D., & Lundquist, J. D. (2015). Estimating mountain basin-mean precipitation from streamflow using Bayesian inference. Water Resources Research, 51(10), 8012–8033. https://doi.org/10.1002/2014WR016736

Henn, B., Painter, T. H., Bormann, K. J., McGurk, B., Flint, A. L., Flint, L. E., et al. (2018). High-Elevation Evapotranspiration Estimates During Drought: Using Streamflow and NASA Airborne Snow Observatory SWE Observations to Close the Upper Tuolumne River Basin Water Balance. Water Resources Research, 54(2), 746–766. https://doi.org/10.1002/2017WR020473

Hogan, D., & Lundquist, J. D. (2024). Recent Upper Colorado River Streamflow Declines Driven by Loss of Spring Precipitation. Geophysical Research Letters, 51(16), e2024GL109826. https://doi.org/10.1029/2024GL109826

#125

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

126 of 178

References (continued)

Hood, E., Williams, M., & Cline, D. (1999). Sublimation from a seasonal snowpack at a continental, mid-latitude alpine site. Hydrological Processes, 13(12–13), 1781–1797. https://doi.org/10.1002/(SICI)1099-1085(199909)13:12/13<1781::AID-HYP860>3.0.CO;2-C

Kampf, S. K., Burges, S. J., Hammond, J. C., Bhaskar, A., Covino, T. P., Eurich, A., et al. (2020). The Case for an Open Water Balance: Re-envisioning Network Design and Data Analysis for a Complex, Uncertain World. Water Resources Research, 56(6), e2019WR026699. https://doi.org/10.1029/2019WR026699

Kapnick, S., & Hall, A. (2012). Causes of recent changes in western North American snowpack. Climate Dynamics, 38(9), 1885–1899. https://doi.org/10.1007/s00382-011-1089-y

Knowles, J. F., Blanken, P. D., Williams, M. W., & Chowanski, K. M. (2012). Energy and surface moisture seasonally limit evaporation and sublimation from snow-free alpine tundra. Agricultural and Forest Meteorology, 157, 106–115. https://doi.org/10.1016/j.agrformet.2012.01.017

Knowles, N., Dettinger, M. D., & Cayan, D. R. (2006). Trends in Snowfall versus Rainfall in the Western United States. Journal of Climate, 19(18), 4545–4559. https://doi.org/10.1175/JCLI3850.1

Lapides, D. A., Hahm, W. J., Rempe, D. M., Whiting, J., & Dralle, D. N. (2022). Causes of Missing Snowmelt Following Drought. Geophysical Research Letters, 49(19), e2022GL100505. https://doi.org/10.1029/2022GL100505

Li, D., Wrzesien, M. L., Durand, M., Adam, J., & Lettenmaier, D. P. (2017). How much runoff originates as snow in the western United States, and how will that change in the future? Geophysical Research Letters, 44(12), 6163–6172. https://doi.org/10.1002/2017GL073551

Lundquist, J. D., & Flint, A. L. (2006). Onset of Snowmelt and Streamflow in 2004 in the Western United States: How Shading May Affect Spring Streamflow Timing in a Warmer World. Journal of Hydrometeorology, 7(6), 1199–1217. https://doi.org/10.1175/JHM539.1

Lundquist, J. D., & Loheide II, S. P. (2011). How evaporative water losses vary between wet and dry water years as a function of elevation in the Sierra Nevada, California, and critical factors for modeling. Water Resources Research, 47(3). https://doi.org/10.1029/2010WR010050

Lundquist, J. D., Cayan, D. R., & Dettinger, M. D. (2004). Spring Onset in the Sierra Nevada: When Is Snowmelt Independent of Elevation? Journal of Hydrometeorology, 5(2), 327–342. https://doi.org/10.1175/1525-7541(2004)005<0327:SOITSN>2.0.CO;2

Lundquist, J. D., Vano, J., Gutmann, E., Hogan, D., Schwat, E., Haugeneder, M., et al. (2024). Sublimation of Snow. Bulletin of the American Meteorological Society, 1(aop). https://doi.org/10.1175/BAMS-D-23-0191.1

Meira Neto, A. A., Niu, G.-Y., Roy, T., Tyler, S., & Troch, P. A. (2020). Interactions between snow cover and evaporation lead to higher sensitivity of streamflow to temperature. Communications Earth & Environment, 1(1), 1–7. https://doi.org/10.1038/s43247-020-00056-9

Milly, P. C. D., & Dunne, K. A. (2020). Colorado River flow dwindles as warming-driven loss of reflective snow energizes evaporation. Science, 367(6483), 1252–1255. https://doi.org/10.1126/science.aay9187

Musselman, K. N., Clark, M. P., Liu, C., Ikeda, K., & Rasmussen, R. (2017). Slower snowmelt in a warmer world. Nature Climate Change, 7(3), 214–219. https://doi.org/10.1038/nclimate3225

Newman, A. J., Clark, M. P., Winstral, A., Marks, D., & Seyfried, M. (2014). The Use of Similarity Concepts to Represent Subgrid Variability in Land Surface Models: Case Study in a Snowmelt-Dominated Watershed. https://doi.org/10.1175/JHM-D-13-038.1

Pagano, T., & Garen, D. (2005). A Recent Increase in Western U.S. Streamflow Variability and Persistence. Journal of Hydrometeorology, 6(2), 173–179. https://doi.org/10.1175/JHM410.1

Pagano, T., Garen, D., & Sorooshian, S. (2004). Evaluation of Official Western U.S. Seasonal Water Supply Outlooks, 1922–2002. Journal of Hydrometeorology, 5(5), 896–909. https://doi.org/10.1175/1525-7541(2004)005<0896:EOOWUS>2.0.CO;2

#126

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

127 of 178

References (continued)

Pagano, T., Wood, A. W., Ramos, M.-H., Cloke, H. L., Pappenberger, F., Clark, M. P., et al. (2014). Challenges of Operational River Forecasting. Journal of Hydrometeorology, 15(4), 1692–1707. https://doi.org/10.1175/JHM-D-13-0188.1

Raleigh, M. S., Livneh, B., Lapo, K., & Lundquist, J. D. (2016). How Does Availability of Meteorological Forcing Data Impact Physically Based Snowpack Simulations? Journal of Hydrometeorology, 17(1), 99–120. https://doi.org/10.1175/JHM-D-14-0235.1

Rasmussen, R., Baker, B., Kochendorfer, J., Meyers, T., Landolt, S., Fischer, A. P., et al. (2012). How Well Are We Measuring Snow: The NOAA/FAA/NCAR Winter Precipitation Test Bed. https://doi.org/10.1175/BAMS-D-11-00052.1

Reba, M. L., Pomeroy, J., Marks, D., & Link, T. E. (2012). Estimating surface sublimation losses from snowpacks in a mountain catchment using eddy covariance and turbulent transfer calculations. Hydrological Processes, 26(24), 3699–3711. https://doi.org/10.1002/hyp.8372

Scaff, L., Krogh, S. A., Musselman, K., Harpold, A., Li, Y., Lillo-Saavedra, M., et al. (2024). The Impacts of Changing Winter Warm Spells on Snow Ablation Over Western North America. Water Resources Research, 60(5), e2023WR034492. https://doi.org/10.1029/2023WR034492

Sexstone, G., Clow, D., Stannard, D. I., & Fassnacht, S. R. (2016). Comparison of methods for quantifying surface sublimation over seasonally snow-covered terrain. Hydrological Processes, 30(19), 3373–3389. https://doi.org/10.1002/hyp.10864

Sexstone, G., Clow, D., Fassnacht, S., Liston, G., Hiemstra, C., Knowles, J., & Penn, C. (2018). Snow Sublimation in Mountain Environments and Its Sensitivity to Forest Disturbance and Climate Warming. Water Resources Research, 54, 1191–1211. https://doi.org/10.1002/2017WR021172

Sturm, M., Goldstein, M. A., & Parr, C. (2017). Water and life from snow: A trillion dollar science question. Water Resources Research, 53(5), 3534–3544. https://doi.org/10.1002/2017WR020840

Svoma, B. M. (2016). Difficulties in Determining Snowpack Sublimation in Complex Terrain at the Macroscale. https://doi.org/10.1155/2016/9695757

Tang, G., Clark, M. P., Knoben, W. J. M., Liu, H., Gharari, S., Arnal, L., et al. (2023). The Impact of Meteorological Forcing Uncertainty on Hydrological Modeling: A Global Analysis of Cryosphere Basins. Water Resources Research, 59(6), e2022WR033767. https://doi.org/10.1029/2022WR033767

Vano, J. A., Das, T., & Lettenmaier, D. P. (2012). Hydrologic Sensitivities of Colorado River Runoff to Changes in Precipitation and Temperature. Journal of Hydrometeorology, 13(3), 932–949. https://doi.org/10.1175/JHM-D-11-069.1

Westerling, A. L., Hidalgo, H. G., Cayan, D. R., & Swetnam, T. W. (2006). Warming and Earlier Spring Increase Western U.S. Forest Wildfire Activity. Science, 313(5789), 940–943. https://doi.org/10.1126/science.1128834

Woodhouse, C. A., & Tintor, W. L. (2024). The Moderating Influence of Spring Climate on the Rio Grande Headwaters: A Paleo Perspective. Water Resources Research, 60(8), e2023WR036909. https://doi.org/10.1029/2023WR036909

#127

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

128 of 178

#128

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

129 of 178

Research objective (illustrated)

Illustrate combining these two event types.

Are problems exacerbated?

Does one abate another? That’s the question.

Determine which have greatest control

How do extremes play a role.

#129

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

130 of 178

#130

Groundwater

Soil Moisture

Groundwater

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

131 of 178

Results so far: NEED TO MAKE OTHER EXAMPLES

#131

HRU 3

Streamflow

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

(b) Model Representation

Develop

Evaluate

Adjust

HRU 6

Elevation Band

HRU 5

HRU 4

Groundwater

Soil Moisture

North Aspect

South Aspect

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

132 of 178

Results so far: NEED TO MAKE OTHER EXAMPLES

#132

(b) Model Representation

Develop

Evaluate

Adjust

HRU 6

Elevation Band

HRU 5

HRU 4

Groundwater

Soil Moisture

North Aspect

South Aspect

Streamflow

Elevation Band

Groundwater

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

133 of 178

Results so far: NEED TO MAKE OTHER EXAMPLES

#133

HRU n

Streamflow

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

…

Groundwater

Soil Moisture

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

134 of 178

1. We’ve got this question of hydrologic anomalies. (See if I can incorporate April forecast for decision making info somewhere?)

2. Many hypotheses to explain them, my masters was devoted to explaining this and we developed some hypotheses

3. We would like a way to test these and we know we need to use a model to represent both time and space

4. Often, we simply choose a model that is built to represent the hydrologic system in a certain way. We mainly have control over parameterizations, knob turning to match our benchmark.

5. But, we now know that model process representation has a lot more to do with how things are connected, what the plumbing is, than the knobs that are turned (Henn 2018)

6. Thus, we’d like to have some control over the model development and figure out which processes are most important to represent.

7. We know that a simple linear model relating SWE to streamflow does a pretty good job, this gives us tons of information and has been the backbone of forecasts for decades.

8. But, we also know that it breaks down in certain years. So, instead we can turn to a fully distributed model to try and represent every process. But again, this can’t do that perfectly, and often these models may do even worse than their counterparts.

9. So is there a sweet spot in between? Where we represent additional processes without going so far as a fully distributed system? This way we can more easily communicate results to those making key decisions.

#134

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

1. We’ve got this question of hydrologic anomalies. (See if I can incorporate April forecast for decision making info somewhere?)

2. Many hypotheses to explain them, my masters was devoted to explaining this and we developed some hypotheses

3. We would like a way to test these and we know we need to use a model to represent both time and space

4. Often, we simply choose a model that is built to represent the hydrologic system in a certain way. We mainly have control over parameterizations, knob turning to match our benchmark.

5. But, we now know that model process representation has a lot more to do with how things are connected, what the plumbing is, than the knobs that are turned (Henn 2018)

6. Thus, we’d like to have some control over the model development and figure out which processes are most important to represent.

7. We know that a simple linear model relating SWE to streamflow does a pretty good job, this gives us tons of information and has been the backbone of forecasts for decades.

8. But, we also know that it breaks down in certain years. So, instead we can turn to a fully distributed model to try and represent every process. But again, this can’t do that perfectly, and often these models may do even worse than their counterparts. (references)

9. So, is there a sweet spot in between? Where we represent additional processes without going so far as a fully distributed system? This way we can more easily communicate results to those making key decisions.

10. What would this look like? Well, maybe a representation of elevation in our basin is needed to model snow accumulation and melt at different times. Or aspect, for the same reasons. Maybe it’s the vegetation type, or how groundwater is connected.

10. How might we evaluate this? We can use SUMMA!

135 of 178

Research objectives – how will we do this?

Well, we're going to use a model to try and build out these different processes in an iterative way, by adding in process complexity piece by piece to see which process representations make the biggest difference against our benchmarks,
while making sure we have our processes grounded in reality using observational networks in heavily instrumented basins.

#135

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Objectives

Well, we're going to use a model to try and build out these different processes in an iterative way, by adding in process complexity piece by piece to see which process representations make the biggest difference against our benchmarks,
while making sure we have our processes grounded in reality using observational networks in heavily instrumented basins.
They have these observations…
We're goin to focus on these three aspects in two different :

Representation of spatial discretization

Hypothesis 1

Representation of subsurface flow

Hypothesis 2

Representation of snow heterogeneity.

Hypothesis 3

How can we represent these differnences.

Center on 2 basins
Well isntrumented (summarize instrumentation with icons)
Focus on these aspects

1,2,3 list. With hypotheses. Add icon representation to illustrate

136 of 178

Methods – Developing modeling alternatives

We have one modeling system that has the capability to build out and represent basin hydrology in different ways. Don't have to worry about different “building materials”.

Is it simple enough to represent with these large blocks
Or do we need these niche components to represent it correctly?

#136

Made by DALL-E

Model Configuration 1 … N

3. Compare model with distributed ET, temperature, and subsurface observations

2. Compare model with streamflow and snow measurement benchmarks

4. Adjust model configuration and parameters to improve HRU connectivity

1. Combine long-term forcing data, land cover data, subsurface data over chosen HRU configuration

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

SUMMA intro

We're gonna use a model called SUMMA.
Why: We're choosing this for a few reasons.

I like to think of SUMMA kind of like legos.
In lego world, we can represent many different systems using the same fundamental components.
SUMMA is kind of similar. Our lego blocks are these things called Hydrologic Response Units (HRU).
We have one modeling system that has the cabability to build out and represent basin hydrology in different ways. Don't have to worry about different building materials.

Is it simple enough to represent with these large blocks
Or do we need these niche compoewnents to represent it correctly?

My analogy isn't perfect, but SUMMA gives us a way to test multiple modeling alternitives to represent the same system and evaluate what makes the biggest difference.
We can run through a cycle of creating a model framework, running it, evaluating it, and adjusting it as needed to try and better represent processes.
In this way, we can test to se which processes are most important to represent well and constrain our efforts?

Introduce SUMMA. Why?

Add lego version of SUMMA.

Build out this analogy.

Add process representation iteration.

Each iteration, we can evaluate its performance against checkpoints

and benchmarks (icons of these)

137 of 178

Methods – Structures for Unifying Multiple Modeling Alternatives (SUMMA)

Why SUMMA: We're choosing this for a few reasons.

I like to think of SUMMA kind of like legos.
In lego world, we can represent many different systems using the same fundamental components.
SUMMA is kind of similar. Our lego blocks are these things called Hydrologic Response Units (HRU).

#137

Made by DALL-E

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

SUMMA intro

We're gonna use a model called SUMMA.
Why: We're choosing this for a few reasons.

I like to think of SUMMA kind of like legos.
In lego world, we can represent many different systems using the same fundamental components.
SUMMA is kind of similar. Our lego blocks are these things called Hydrologic Response Units (HRU).
We have one modeling system that has the cabability to build out and represent basin hydrology in different ways. Don't have to worry about different building materials.

Is it simple enough to represent with these large blocks
Or do we need these niche compoewnents to represent it correctly?

My analogy isn't perfect, but SUMMA gives us a way to test multiple modeling alternitives to represent the same system and evaluate what makes the biggest difference.
We can run through a cycle of creating a model framework, running it, evaluating it, and adjusting it as needed to try and better represent processes.
In this way, we can test to se which processes are most important to represent well and constrain our efforts?

Introduce SUMMA. Why?

Add lego version of SUMMA.

Build out this analogy.

Add process representation iteration.

Each iteration, we can evaluate its performance against checkpoints

and benchmarks (icons of these)

138 of 178

Transition seasons have outsized affect on streamflow

These transition seasons have an outsized effect that has been recognized more widely in recent years.
Changes in spring and fall have been identified as particularly important for controlling runoff.

Fall soil moisture
Spring precipitation

but these physical processes responsible for these disproportionate effects are not well understood.
Annual changes are often applied, but not so much specific seasons, outside of winter

#138

Cite relevant work, screenshots. Maybe show a plot of water year time series?

Highlight periods of interest.

Physical processes are not well constrained.

This isolates effects on the seasonal scale.

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

139 of 178

Recent hydrologic anomalies attributed to transition season conditions

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in response. More loss of water during spring. Exacerbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

#139

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

What are these impacts? What are the hypotehses?

Initial conditions set up springtime melt period.
Fall soil moisture, dry antecedent conditions are blamed. Soil "bucket" must refill before runoff. Lower runoff efficiency.

Winter Tsoil and θ were both higher when early winter snow accumulation was greater (Maurer and Bowling 2014)

Of particular importance are changes in fall and early winter snowpack development, as seasonal snowpacks isolate the soil environment until spring snowpack ablation begins.

Warm, dry springs induce earlier snowmelt. Plants to turn on earlier in resposne. More loss of water during spring. Excaberbated effect from lost precip.

Soil moisture availability controlled by when ground is exposed. Earlier melt out, earlier heating at effect

Wet, warm conditions mean more rain and less snow from this is measured but not well understood why. Conflicting answers.

List hypotheses with illustration. Show simple model representation.

Dry, warm fall vs wet, cool fall:

Dry soil, wet soil.
Add snow, insulator for each
Dry soil maintained until later, wet soil sets up.
Snow melts, but soil is dry, water goes to soil and ET, then runoff. Lower runoff efficiency. Soil water promotes runoff.

Warm dry spring vs. wet, cool spring

Normal soil in each
Snow in each.
Dry, sunny warm spring. Cool wet spring
Smaller snowpack, melts earlier, earlier ET, drier soils in summer. Later snowmelt, faster snowmelt, less early spring ET, integrates to smaller amount in spring. Changes runoff efficiency and timing of tunoff.

Highlight papers that focus on specific changes over periods relatively short periods of the season. Big storms, big warm periods.

140 of 178

Research objectives

Uncertainty in combined effects. Do they build off each other? Do they cross each other out? Is one more important than another?
Our research goals will focus on these three areas:

Fall conditions -> warm + wet, cold + dry, warm + dry, cold + wet
Spring conditions -> warm + wet, cold + dry, warm + dry, cold + wet
Large events that control seasonal output.

We'll look to see what has a greater control: initial conditions set up in fall or springtime conditions?
Also investigate impact of season averages controlled by short duration extreme events

#140

Chapter 2 🌧️ (ongoing)

Chapter 3 🏔️ (ongoing)

Chapter 4 🍂/🌱 (proposed)

Chapter 1

Conclusions & �Timeline

Introduction

141 of 178

Uncertainties can be constrained by measurements

#141

Eddy covariance system �(PC: Emilio Mateo)

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Difficult to model in complex terrain

Its difficult to model because its representation relies on assumptions that are commonly invalidated in complex terrain. I won't go over these reasons, but here but Eli likely has some thoughts on the subject. These can break down in certain conditions common to these mountain environments.

Rarely measured in complex terrain. Measurement is also tricky. There are a few different methods used to estimate it in the field, but only really one way to measure it.
This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Not considered in the current USGS water cycle.

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

142 of 178

Uncertainties can be constrained by measurements

But measurements face challenges

Difficult to take and maintain,
Expensive

#142

Eddy covariance system �(PC: Emilio Mateo)

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Difficult to model in complex terrain

Its difficult to model because its representation relies on assumptions that are commonly invalidated in complex terrain. I won't go over these reasons, but here but Eli likely has some thoughts on the subject. These can break down in certain conditions common to these mountain environments.

Rarely measured in complex terrain. Measurement is also tricky. There are a few different methods used to estimate it in the field, but only really one way to measure it.
This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Not considered in the current USGS water cycle.

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

143 of 178

Uncertainties can be constrained by measurements

But measurements face challenges

Difficult to take and maintain,
Expensive

Thus, past observations were often isolated to short observation periods

#143

Dec 1

Cumulative Sublimation

Mar 31

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Difficult to model in complex terrain

Its difficult to model because its representation relies on assumptions that are commonly invalidated in complex terrain. I won't go over these reasons, but here but Eli likely has some thoughts on the subject. These can break down in certain conditions common to these mountain environments.

Rarely measured in complex terrain. Measurement is also tricky. There are a few different methods used to estimate it in the field, but only really one way to measure it.
This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Not considered in the current USGS water cycle.

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

144 of 178

Uncertainties can be constrained by measurements

But measurements face challenges

Difficult to take and maintain, expensive

Thus, past observations were often isolated to short observation periods

#144

Dec 1

Cumulative Sublimation

Mar 31

So, what can we do?

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Difficult to model in complex terrain

Its difficult to model because its representation relies on assumptions that are commonly invalidated in complex terrain. I won't go over these reasons, but here but Eli likely has some thoughts on the subject. These can break down in certain conditions common to these mountain environments.

Rarely measured in complex terrain. Measurement is also tricky. There are a few different methods used to estimate it in the field, but only really one way to measure it.
This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Not considered in the current USGS water cycle.

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

145 of 178

Comparing event types at site-scale: surface energy balance

#145

Plan to build these plots out and change colors!!!

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

First, sublimation is energy intensive. We can look to see if the source of energy is different during these two event types.

We see that it is.

We can look at temperature.

We see that it is

We can look at humidity.

We see that it is

We can look at wind speed.

We see that it is.

Thus at the site-scale. These events have different characteristics.
Now we can scale up to the valley.

We can use these observations to see that its also different between event types.

Now scale up to synoptic.

Short, intense events likely due to ridge overhead.
Long events related to persistent, windy conditions aloft.

But trajectory of storms has huge implications!

Go back to observation diagram. Highlight site-scale

Then switch to figure comparing years. Build it sequentially by introducing things one at a time

Add in temperature. Discuss differences

Add in humidity discuss differences

Add in wind speed, discuss differences.

Takaway.

Go back to observation diagram. Highlight valley-scale. What processes differ. Wintertime inversion stength + turbulence/windy conditions throughout the valley.

Show results for both.

Go back to observation diagram. Highlight synoptic scale.

Show average conditions over the WUS during different events. Maybe average ERA5 results over the basin? Show average temperature, wind speed, humidity during events?

Illustrate final steps. Use this idea as a lens to seek at large events in the past

146 of 178

Comparing event types at site-scale: temperature

#146

Plan to build these plots out and change colors!!!

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

First, sublimation is energy intensive. We can look to see if the source of energy is different during these two event types.

We see that it is.

We can look at temperature.

We see that it is

We can look at humidity.

We see that it is

We can look at wind speed.

We see that it is.

Thus at the site-scale. These events have different characteristics.
Now we can scale up to the valley.

We can use these observations to see that its also different between event types.

Now scale up to synoptic.

Short, intense events likely due to ridge overhead.
Long events related to persistent, windy conditions aloft.

But trajectory of storms has huge implications!

Go back to observation diagram. Highlight site-scale

Then switch to figure comparing years. Build it sequentially by introducing things one at a time

Add in temperature. Discuss differences

Add in humidity discuss differences

Add in wind speed, discuss differences.

Takaway.

Go back to observation diagram. Highlight valley-scale. What processes differ. Wintertime inversion stength + turbulence/windy conditions throughout the valley.

Show results for both.

Go back to observation diagram. Highlight synoptic scale.

Show average conditions over the WUS during different events. Maybe average ERA5 results over the basin? Show average temperature, wind speed, humidity during events?

Illustrate final steps. Use this idea as a lens to seek at large events in the past

147 of 178

Comparing event types at site-scale: wind speed

#147

Plan to build these plots out and change colors!!!

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

First, sublimation is energy intensive. We can look to see if the source of energy is different during these two event types.

We see that it is.

We can look at temperature.

We see that it is

We can look at humidity.

We see that it is

We can look at wind speed.

We see that it is.

Thus at the site-scale. These events have different characteristics.
Now we can scale up to the valley.

We can use these observations to see that its also different between event types.

Now scale up to synoptic.

Short, intense events likely due to ridge overhead.
Long events related to persistent, windy conditions aloft.

But trajectory of storms has huge implications!

Go back to observation diagram. Highlight site-scale

Then switch to figure comparing years. Build it sequentially by introducing things one at a time

Add in temperature. Discuss differences

Add in humidity discuss differences

Add in wind speed, discuss differences.

Takaway.

Go back to observation diagram. Highlight valley-scale. What processes differ. Wintertime inversion stength + turbulence/windy conditions throughout the valley.

Show results for both.

Go back to observation diagram. Highlight synoptic scale.

Show average conditions over the WUS during different events. Maybe average ERA5 results over the basin? Show average temperature, wind speed, humidity during events?

Illustrate final steps. Use this idea as a lens to seek at large events in the past

148 of 178

Comparing event types at site-scale: vapor pressure deficit

#148

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

First, sublimation is energy intensive. We can look to see if the source of energy is different during these two event types.

We see that it is.

We can look at temperature.

We see that it is

We can look at humidity.

We see that it is

We can look at wind speed.

We see that it is.

Thus at the site-scale. These events have different characteristics.
Now we can scale up to the valley.

We can use these observations to see that its also different between event types.

Now scale up to synoptic.

Short, intense events likely due to ridge overhead.
Long events related to persistent, windy conditions aloft.

But trajectory of storms has huge implications!

Go back to observation diagram. Highlight site-scale

Then switch to figure comparing years. Build it sequentially by introducing things one at a time

Add in temperature. Discuss differences

Add in humidity discuss differences

Add in wind speed, discuss differences.

Takaway.

Go back to observation diagram. Highlight valley-scale. What processes differ. Wintertime inversion stength + turbulence/windy conditions throughout the valley.

Show results for both.

Go back to observation diagram. Highlight synoptic scale.

Show average conditions over the WUS during different events. Maybe average ERA5 results over the basin? Show average temperature, wind speed, humidity during events?

Illustrate final steps. Use this idea as a lens to seek at large events in the past

149 of 178

At the site-scale, events have different characteristics

#149

Plan to build these plots out and change colors!!!

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

First, sublimation is energy intensive. We can look to see if the source of energy is different during these two event types.

We see that it is.

We can look at temperature.

We see that it is

We can look at humidity.

We see that it is

We can look at wind speed.

We see that it is.

Thus at the site-scale. These events have different characteristics.
Now we can scale up to the valley.

We can use these observations to see that its also different between event types.

Now scale up to synoptic.

Short, intense events likely due to ridge overhead.
Long events related to persistent, windy conditions aloft.

But trajectory of storms has huge implications!

Go back to observation diagram. Highlight site-scale

Then switch to figure comparing years. Build it sequentially by introducing things one at a time

Add in temperature. Discuss differences

Add in humidity discuss differences

Add in wind speed, discuss differences.

Takaway.

Go back to observation diagram. Highlight valley-scale. What processes differ. Wintertime inversion stength + turbulence/windy conditions throughout the valley.

Show results for both.

Go back to observation diagram. Highlight synoptic scale.

Show average conditions over the WUS during different events. Maybe average ERA5 results over the basin? Show average temperature, wind speed, humidity during events?

Illustrate final steps. Use this idea as a lens to seek at large events in the past

150 of 178

Valley-scale observations: Temperature inversion intensity & depth

#150

T1

T2>T1

T3>T2

T4<T3

Inversion -> temperature increases with height

Inversion Depth: e.g. Z3 – Z0�Inversion Intensity: Temperature change (e.g. T3 – T1)

Z3

Z2

Z1

Z0

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

151 of 178

Valley-scale observations: Temperature inversion intensity & depth

#151

Weak inversions during long events. Inversion breaks down during afternoon.

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

152 of 178

Valley-scale observations: vertical velocity variance

#152

Calculated from doppler lidar vertical staring periods�Serves as measure of turbulence aloft

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

153 of 178

Valley-scale observations: Vertical Velocity Variance

#153

Greater connection between conditions aloft and at surface during events

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

154 of 178

Background – Types of observation + uncertainties

#154

Accumulation gauges

Gauge undercatch
Snow bridging
Maintenance�Precip type

Weight-based Gauges

Gauge undercatch
Snow bridging
Maintenance

Optical gauges�

Maintenance, power requirements
Can’t differentiate blowing from falling precip
Also reliant on rain rate relationships to particles

Radar

Only on that can give us distributed estimates
Beam blockage
Relies on Z-R, Z-S relationship
Sensitive to precip type

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

Why is it so uncertain?

Observations sound pretty simple. Go to mountains and catch or measure all the precip that falls over a location. That doesn't make it easy.

Questions about how we scale obs?

Its done a few different ways, each exposed to their own errors

Optical gauges - list

Discuss how precip is calculated for disdrometers.

Accumulation gauges - list
Weight-based gauges - list
Radar (only on that can give us distributed estimates)

Surface or satellite based. Satellite are unreliable in compelx terrain
Rely on Z-R relationship. Highly sensitive to precip type.

Uncertainties can be adjusted, usually through adjustments , but unaccounted for uncertainties can propagate through gridded datasets.
Can cause issues with interpreting relationships b/w precip and other results.

Start with simple representation of measurement.

Show picture of different observation types.

Optical -> list, discuss calculation + uncertainties
Accumulation gauges -> list, discuss calculation + uncertainties
Weight based -> list, discuss calculation + uncertainties
Radar -> discuss calc + uncertainties

References to adjustments.

Show figure of integrating gauge observations into grid via interpolation.

Flaws in measurement propagate.

155 of 178

155

As with the model decisions, the column specification is set globally for the entire model domain. The

specific column configuration can be selected in the model decisions file. Whether the columns contribute

to a single or common aquifer per GRU is indicated by the choice of the spatial_gw decision in that file.

Spatial exchange between HRUs is indicated by specifying the downslope HRU in the local attributes file.

156 of 178

Background – Models are used to represent these “balances”

#156

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

157 of 178

Motivation – Snow-dominated catchments have been changing

Several changes have been observed in recent decades in response to:

temperature rise,
changing precipitation patterns,
natural variability

#157

Earlier snowmelt

Earlier peak SWE timing

Less snow, more rain

Earlier spring onset

More frequent wildfires

Streamflow decline

Abatzoglou & Williams (2016)

Westerling et al. (2006)

Cayan et al. (2001)

Westerling et al. (2006)

Clow (2010)

McCabe & Clark (2005)

Kapnick & Hall (2012)

Hamlet et al. (2005)

Knowles et al. (2006)

Musselman et al. (2017)

Hogan & Lundquist (2024)

Early Normal Late

SWE

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

158 of 178

158

159 of 178

159

February 9, 2025, HWY 97 S of Entiat, WA

February 9, 2025, HWY 97 NE of Chelan, WA

August 8, 2022 S of White Mtn, Glacier Peak Wilderness

160 of 178

160

161 of 178

161

162 of 178

162

163 of 178

163

0. Use precipitation gauge observations (Chapter 2) and

Model Configuration 1 … N

3. Compare model with distributed ET, temperature, and subsurface observations

2. Compare model with streamflow and snow measurement benchmarks

4. Adjust model configuration and parameters to improve HRU connectivity

1. Combine long-term forcing data, land cover data, subsurface data over chosen HRU configuration

164 of 178

164

Model Configuration 1 … N

3. Compare model with distributed ET, temperature, and subsurface observations

2. Compare model with streamflow and snow measurement benchmarks

1. Combine long-term forcing data, land cover data, subsurface data over chosen HRU configuration

165 of 178

165

HRU 3

Streamflow

̄P = 350 mm

_

̄P = 275 mm

_

̄P = 200 mm

_

Elevation Band

HRU 2

HRU 1

Groundwater

Soil Moisture

(a) Forcing Data Applied to HRUs

(b) Model Representation

(c) Model Output Comparison

Adjust model configuration and parameters to improve HRU connectivity. Repeat…

166 of 178

166

167 of 178

167

168 of 178

Cooler/Wetter

Warmer/Drier

(b) Tuolumne River Basin

(a) East River Basin

Standardized Anomaly

169 of 178

169

170 of 178

What drives uncertainty in sublimation?

Sublimation estimates range from 1% to 90% of seasonal snowfall locally �(Lundquist et al., 2024; Svoma et al., 2016)
Sublimation increases often linked directly to temperature rise �(Scaff et al., 2024; Harpold et al., 2012; Harpold & Brooks, 2018)
But observations show periodic events are a more important signal�(Bliss et al., 2011; Hood et al. 1999; Knowles et al., 2012, Reba et al, 2012)

#170

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

This contributes to uncertainty around the dominant processes that control it.

There is uncertainty around the dominant processes that drive seasonal sublimation. Some areas have shown short events are important, but there are few investigations of whole seasons of observation, let alone multiple years. Those that do explore full seasons normally just report what the seasonal totals were, not the dominant contributors to those totals.
Does sublimation behave more like big storms, where short periods cause lots of it or does it build up sequentially over the winter?
Blowing snow sublimation is important. What's its role. Blowing snow is often not measured.
The route that it takes is important to understand because our models might represent one well and another poorly?

Reference huge variability in models. Causes uncertainty in streamflow outputs.

Show image of EC setup in complex terrain.

Reference papers that talk about dominant processes.

Show blowing snow images from MacDonald.

Create cartoon depiction of sublimation build up:

Consistent line
Piecewise build up.

171 of 178

171

172 of 178

172

173 of 178

173

174 of 178

174

Sublimation/Evaporation

Cumulative Sublimation/Evaporation (mm)

Time

175 of 178

175

176 of 178

Motivation – Snow-dominated catchments have been changing

Several changes have been observed in recent decades in response to:

temperature rise,
changing precipitation patterns,
natural variability

Earlier snowmelt

Earlier peak SWE timing

Less snow, more rain

Earlier spring onset

Streamflow decline

Cayan et al. (2001)

Westerling et al. (2006)

Clow (2010)

McCabe & Clark (2005)

Kapnick & Hall (2012)

Hamlet et al. (2005)

Knowles et al. (2006)

Musselman et al. (2017)

Hogan & Lundquist (2024)

Early Normal Late

SWE

#176

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

177 of 178

Motivation – Streamflow response to these changes still vary

Streamflow response to these change varies due to differences in:

regional runoff efficiency �(Harpold et al., 2012; Brooks et al. 2021)
hydrologic process representation in models �(Bennett et al., 2019)
sensitivity of streamflow to temperature rise & precipitation change �(Vano et al., 2012)

#177

SWE

Chapter 2 🌧️ (ongoing)

Chapter 1

Conclusions & �Timeline

Introduction

Chapters 3 🏔️ & 4 🍂/🌱 �(proposed)

178 of 178

178