UC San Diego

This dissertation demonstrates the uses of satellite and surface observations, in tandem with hydrologic modeling, to characterize daily-to-interannual cloudiness variability and its influence on spring-summer snowmelt and streamflow fluctuations over the mountains of the western United States from 1996 to 2015.

Daily cloudiness variations can exceed 50% of long-term averages during the springtime. When aggregated over three-month periods, cloudiness varies by ±10% of long-term averages in many locations. Rotated empirical orthogonal functions (REOFs) analysis indicates the first five REOFs account for ~67% of the total variance, each of which has distinct regional and seasonal emphases. Each of the REOF modes associates with anomalous large scale atmospheric circulation patterns and one or more large-scale teleconnection indices, which helps to explain why anomalous cloudiness patterns take on regional spatial scales and contain substantial variability over seasonal time scales.

Cloud cover indices (CC) are, to some extent, related linearly to snowmelt (ΔSWE) and snow-fed streamflow (ΔQ) fluctuations. Local CC-ΔSWE and CC-ΔQ associations vary with time and location, with the dominance of negative correlations between CC and ΔSWE, exemplifying the cloud-shading (or clear-sky) effect on snowmelt. The magnitude of CC-ΔSWE association (R2) amounts to 5%-56%, typically peaking in May. These associations fade earlier in summer during dry years than wet years, indicating the differing responses of higher vs. lower snowpack. The CC-ΔQ association displays less consistent arrangement, with R2 amounting to 2%-47%. The ΔSWE and ΔQ fluctuations exhibit spatially extensive patterns of correlations with daily CC anomalies, indicating the effects of cloudiness often operate over regional scales.

On a watershed scale, cloudiness variability redistributes the seasonal runoff and hastens the spring onset by 1-3 days. Higher elevation cloudiness exerts a greater influence on the basin runoff than lower elevation cloudiness does. Overall, cloudiness delays spring onset by 2-15 days regardless of the elevation. Lastly, the experiment on the intensification of cloudiness fluctuations suggests greater streamflow sensitivity to the “relatively cloudy periods get cloudier” scheme than to the “relatively clear periods get clearer” scheme, with the former producing 3-5 days later spring onsets.