1 of 15

An Overview of Satellite Precipitation Products

Jackson Tan

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

University of Maryland, Baltimore County & NASA Goddard Space Flight Center

jackson.tan@nasa.gov

2 of 15

From Satellites to Precipitation Estimates

  • Retrieval process:
    • Retrieval method: active vs. passive
    • Sensor frequency: MW vs. IR/VIS
    • Satellite platform: LEO vs. GEO
    • Data processing techniques
  • Merging the satellite observations
  • Selecting a precipitation product

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

3 of 15

Retrieval Method: Active vs. Passive

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

Generally more accurate and provide more information. Examples: GPM DPR, TRMM PR, CloudSat CPR.

Wider swaths and more instruments in orbit. Examples: GPM GMI, NOAA MHS, GOES ABI

Adapted from https://svs.gsfc.nasa.gov/11877

4 of 15

Sensor Frequency: MW, IR, and VIS

  • Electromagnetic radiation is emitted by any object with a temperature above absolute zero at a range of wavelengths that depends on its temperature and emissivity.
  • Electromagnetic radiation is scattered by particles with diameters similar to or larger than the wavelength of the radiation.
  • Microwave (MW): scattered by precipitation particles but not cloud droplets
    • But precipitation particles can also emit MW radiation, so whether a retrieval algorithm relies on the emission or scattering signal depends on the “brightness” (or emissivity) of the surface.
    • In general, precipitation retrievals rely on the emission signal of the precipitation over ocean at low MW frequencies and the scattering signal of the precipitation over land at high MW frequencies.
  • Infrared (IR) & Visible (VIS): scattered by cloud particles, use information about the clouds to infer the precipitation
    • VIS: sunlit hours only

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

https://science.nasa.gov/ems/01_intro

cloud droplets

precipitation droplets

Sun’s emitted emission

Earth’s emitted emission

5 of 15

Satellite Orbit: LEO vs. GEO

  • Low-Earth orbit (LEO), being closer to the Earth’s surface, provides better resolution.
  • Geosynchronous orbit (GEO) is at a vantage point that allows constant monitoring.
  • Since footprint resolution depends on wavelength, the orbit is usually limited by the sensor frequency.
  • VIS and IR sensors are on both GEO and LEO platforms, with resolutions reaching O(1 km) and O(100 m) respectively.
  • MW sensors are only on LEO, with resolutions reaching O(1 km) but more often in O(10 km).
    • Antenna configuration can increase resolution, but this involves engineering challenges.

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

6 of 15

Data Processing Techniques

  • Satellites do not measure precipitation. Satellites measure electrical signals or radiation.
  • From satellite measurements to precipitation products:
    • Algorithms to convert the satellite measurements to precipitation rates.
    • Calibration between different sensors.
    • Combining observations from different satellites.
    • Filling in gaps and quality control.
  • Choices above is guided by design priority: what does the product prioritize?
    • A consistent long-term record (Climate Data Record) or the best estimate at any one time (High Resolution Precipitation Product)?

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

7 of 15

Data Processing Levels

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

Data Level

Description

Level 0

Raw instrument data

Level 1A

Reconstructed, unprocessed instrument data at full resolution, time referenced, and annotated with ancillary information, including radiometric and geometric calibration coefficients and georeferencing parameters (i.e., platform ephemeris), computed and appended, but not applied, to Level 0 data.

Level 1B

Radiometrically corrected and geolocated Level 1A data that have been processed to sensor units.

Level 2

Derived geophysical parameters at the same resolution and location as those of the Level 1 data.

Level 3

Geophysical parameters that have been spatially and/or temporally resampled from Level 1 or Level 2 data.

Level 4

Outputs or results from models using lower level data as inputs and, thus, not directly derived from the instruments.

https://gpm.nasa.gov/data-access/data-products

Brightness temperature

Precipitation at satellite footprints

Gridded precipitation

Satellite + model precipitation

8 of 15

Using IMERG as Example of Merging

  • We have a collection of separate observations from different satellites at various frequencies and orbits, but users want a consistent, gridded precipitation estimate at regular intervals.
  • The Integrated Multi-satellitE Retrievals for GPM (IMERG): gridded precipitation product from the U.S. Science Team from the GPM Mission.
    • Grid the MW precipitation estimates to 0.1° every half-hour and calibrate them to the combined DPR and GMI product.
    • Apply morphing: interpolation by motion and IR precipitation.
    • Perform monthly gauge adjustment.
  • Other gridded satellite products are produced using variations or subsets of these steps.

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

9 of 15

Intercalibrating the MW Precipitation

  • The fact that the different satellite sensors have different characteristics mean that their precipitation estimates may have different distributions, even if they use calibrated inputs and a consistent algorithm.
  • We want to calibrate the MW precipitation estimates from all satellites to the best satellite precipitation estimates provided by the GPM combined DPR and GMI product.
  • Calibration is a delicate balance between a long record (enough sample size) and a short record (representing “current” conditions).

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

10 of 15

Morphing: Core Concept

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

08:30

10:00

09:00

09:30

11 of 15

Morphing: Combining Interpolated Precip.

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

+

=

Propagated precipitation

IR precipitation

12 of 15

Which Satellite Product to Use?

  • What are you using the satellite product for? What are the requirements you have of the satellite product?
    • Resolution, coverage, latency?
    • What sort of precipitation observations do you need? How accurate do they need to be?
    • Climate Data Record?
    • Observations combined with models?
  • Non-exhaustive list of satellite precipitation products:
    • https://gpm.nasa.gov/data/directory
    • http://ipwg.isac.cnr.it/data/datasets.html
    • https://climatedataguide.ucar.edu/climate-data/precipitation-data-sets-overview-comparison-table
  • Data provider vs. third-party host

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

13 of 15

Satellite Precipitation vs. Others

  • Compared to field experiments, satellite precipitation has:
    • wider coverage and longer record;
    • the configuration to provide timely data with low latency; and
    • lower accuracy and resolution, more limited information.
  • Compared to models, satellite precipitation (currently) has:
    • generally better resolution + coverage;
    • generally higher accuracy in convective environments, similar or lower accuracy in stratiform environments;
    • shorter record; and
    • no forecast data.

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

14 of 15

Summary

  • Retrieval process:
    • Retrieval method: active vs. passive → information content
    • Sensor frequency: MW vs. IR/VIS → precipitation or cloud
    • Satellite platform: LEO vs. GEO → resolution and sampling
    • Data processing techniques → from measurement to precipitation
  • Merging the satellite observations → methods of combining observations
  • Selecting a precipitation product → requirements, pros & cons over field experiments and models

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022

15 of 15

Useful Resources

  • Product listings:
    • GPM website: https://gpm.nasa.gov/data/directory
    • IPWG: http://ipwg.isac.cnr.it/data/datasets.html
    • UCAR: https://climatedataguide.ucar.edu/climate-data/precipitation-data-sets-overview-comparison-table
  • Review articles on satellite precipitation products:
    • Kidd and Huffman (2011): 10.1002/met.284
    • Kidd and Levizzani (2011): 10.5194/hess-15-1109-2011
    • Tapiador et al. (2012): 10.1016/j.atmosres.2011.10.021
    • Kidd et al. (2021): 10.1175/BAMS-D-20-0299.1

AGU Precipitation Technical Committee Quarterly Seminar | 20th Sep 2022