Exploring the Role of Soil Moisture in Seasonal Streamflow Forecast over the Upper Colorado River Basin
Abstract
The Colorado River is well-known for heavily regulated by major reservoirs to provide water supply for the arid southwest United States. Although the area of the Upper Basin (defined as upstream part above Lees Ferry) and that of the Lower Basin are equivalent, approximately 90% of the natural flow in the river is generated in the upper part. The majority of the natural flow in the Upper Basin originates from liquid water generated by snowmelt, and the value of annual peak Snow Water Equivalent (SWE) to seasonal streamflow prediction has been confirmed by the hydrology community. However, the role of pre-season soil moisture (soil moisture storage before the snow season) in spring runoff generation remains underexplored. Based on a well-established surface hydrological simulation in the Upper Basin, we employed random forest regression method to evaluate the contribution of pre-season soil moisture to next year's Apr-Jul runoff prediction. Although peak SWE is typically the most significant contributor to streamflow forecast, the role of initial soil moisture is also very important. The regression analysis demonstrates that >70% of the total natural flow in Apr-Jul is more dependent on soil moisture anomaly than on Apr-Jul precipitation, especially in the region with high runoff productivity. The impact of soil moisture to streamflow prediction is greater than that of spring precipitation until July. Our results indicates that the pre-season soil moisture can provide good insights to streamflow prediction and have the potential to improve seasonal forecast skills.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H22F..03X