Effects of Antecedent Soil Moisture on Surface Runoff During Storm Events

1^st Tarrik Quneibi¹

¹Civil and Environmental Engineering, University of Michigan

​

Introduction

Antecedent soil moisture refers to the relative wetness in the upper soil layer. In literature, antecedent conditions have been seen to play a major role in flow responses during storm events, although more commonly in areas with highly permeable surfaces such as in grasslands and forests. In studies, it has been seen that antecedent soil moisture does not play as major of a role when there are impervious surfaces. Antecedent soil moisture is a major factor when modelling sanitary sewers as the conditions can lead to higher runoff to accumulate which will require a large sanitary sewer. To add to the complexity, antecedent soil moisture is not only affected by preceding storm events, but is also dependent on air temperature, wind speed, humidity levels, evapotranspiration, and snow and ice melting.

​

​

​

Objectives

1. To compare data from field sensors to soil moisture satellite data to determine the accuracy of the satellite data
2. To determine the correlation (if any) between storm events and the increase/decrease of soil moisture in the subsurface.
3. To determine how wet and dry soil conditions affect the discharge into rivers.

Materials and Methods

Two locations, one urban and one rural, were located close to the USGS discharge sensor so that an accurate representation of runoff could be obtained. The urban location was chosen to be WSU woodward/Harper parking lot in Detroit, while the “rural” location was The Huron river at Maiden Lane in Ann Arbor. Field sensor data was pulled for these two locations. This data included Subsurface water depth, surface water depth, and discharge. Satellite data was then pulled from Google Earth Engine API using Python 3 along with the coordinates of the selected locations. This data included percent soil moisture (%), surface moisture depth (mm), subsurface moisture depth (mm) from NASA-USDA Enhanced SMAP (resolution: 10km), and gage corrected precipitation (mm/hr) from GSMaP Operational (resolution: 11.13km) . All data were then imported into R studio to be cleaned and analyzed. Once all data was cleaned, the field sensor subsurface water level was compared to the satellite subsurface data to determine the accuracy of the satellite data. Data for precipitation, soil moisture, subsurface water depth, and discharge were then plotted to determine correlations between these parameters. All figures were created using the ggplot package.

​

Results/Modeling

Observation 1: The baseline subsurface water depth increases/decreases along with antecedent soil moisture.

Discussion and Conclusions

• Since the resolution for the antecedent soil moisture data is 10km, it does not contain the same peaks that the field sensor has. However, it does follow similar baseline trends (field data and satellite data increase/decrease in the same intervals).
• The antecedent soil moisture data showed very little difference, if any, between the “rural” and urban locations. This may be due to the resolution of the data, or possibly that the ground surface of both locations are similar in permeability. Since both locations experience similar air temperature, wind speed, storm events, etc, permeability of the surface is likely a larger factor.
• From Figure 2 we can see that each storm events increases the subsurface water depth and corresponds to each precipitation peak. However, the antecedent soil moisture seems to increase over time with more frequency of storms rather than just the magnitude of the storm. This could be attributed to other parameters such as humidity and air temperature causing evapotranspiration between storm events. If the frequency of storms is higher, then the subsurface has less time to dry.
• The discharge in the river peaks along with major storm events, which is to be expected. Furthermore, the baseline discharge increases along with the antecedent soil moisture. In the winter months, the soil moisture is almost 100% wet due to snow melt and freezing. Through these months are where we see the highest discharge values as well as the highest average baseline.
• Further analysis is needed before concluding any major correlation between antecedent soil moisture and surface runoff. The other parameters listed in the introduction must also be considered.

​

Acknowledgements

Google Earth Engine API

Professor Kerkez

References

Observation 2: Storm events cause increase in subsurface water depth, but antecedent soil moisture depends on frequency of events.

Observation 3: As antecedent soil moisture increases, the discharge also increases.

​

[1] “Soil Moisture Profiles and Temperature Data from SoilSCAPE Sites, USA” (2017)

[2] Hirokazu (2005)

[3] Schoener 627-636 (2019)

Figure 1. Satellite antecedent soil moisture over time aligned with field sensor subsurface water depth. When broken up into sections of low and high antecedent soil moisture, the baseline subsurface water depth (shown in green) raises and lowers with the soil moisture percent. As satellite data has a spatial resolution of 10km it is not as accurate to the specific coordinate as the field sensor, but does follow a similar trend.

Figure 2. Subsurface water depth, precipitation, and antecedent soil moisture over time. The storm event peaks correspond fairly well to the subsurface water depth, but the antecedent soil moisture seems to increase more based on the frequency of storm events rather than the magnitude.

Figure 3. The antecedent soil moisture appears to show seasonal variation which is expected due to storm seasons and snow melt. As the percent soil moisture increases, the discharge increases. The discharge peaks correspond to the precipitation peaks, but the baseline of the discharge increases/decreases with antecedent soil moisture.

Observation 1: The baseline subsurface water depth increases/decreases along with antecedent soil moisture.

Figure 1. Satellite antecedent soil moisture over time aligned with field sensor subsurface water depth. When broken up into sections of low and high antecedent soil moisture, the baseline subsurface water depth (shown in green) raises and lowers with the soil moisture percent. As satellite data has a spatial resolution of 10km it is not as accurate to the specific coordinate as the field sensor, but does follow a similar trend.

Effects of Antecedent Soil Moisture on Surface Runoff During Storm Events

Effects of Antecedent Soil Moisture on Surface Runoff During Storm Events

Observation 2: Storm events cause increase in subsurface water depth, but antecedent soil moisture depends on frequency of events.

Figure 2. Subsurface water depth, precipitation, and antecedent soil moisture over time. The storm event peaks correspond fairly well to the subsurface water depth, but the antecedent soil moisture seems to increase more based on the frequency of storm events rather than the magnitude.

Effects of Antecedent Soil Moisture on Surface Runoff During Storm Events

Observation 3: As antecedent soil moisture increases, the discharge also increases.

​

Figure 3. The antecedent soil moisture appears to show seasonal variation which is expected due to storm seasons and snow melt. As the percent soil moisture increases, the discharge increases. The discharge peaks correspond to the precipitation peaks, but the baseline of the discharge increases/decreases with antecedent soil moisture.