Bioscience Evidence 2024, Vol.14, No.4, 161-171 http://bioscipublisher.com/index.php/be 166 Moreover, models that blend different data sources, such as precipitation gauge observations with snowpack measurements, have shown improved predictions of snowpack evolution. In the Kings River Basin, California, a blended scenario combining these data sources resulted in better snowpack predictions compared to models relying solely on precipitation gauges or standard gridded products (Figure 2) (Pelak et al., 2022). 6.3 Integration of remote sensing and ground-based observations The integration of remote sensing and ground-based observations has enhanced the monitoring and modeling of snowpack. Remote sensing technologies, such as the Moderate Resolution Imaging Spectrometer (MODIS), provide valuable data on snow cover, which can be used to supplement sparse in situ measurements in data-poor regions (Sproles et al., 2016). Additionally, the assimilation of passive microwave satellite observations, such as those from the AMSR-2, into snowpack models has improved SWE estimates by updating meteorological forcing data and snowpack states without relying on surface-based data (Larue et al., 2018). Figure 2 (a, b) Box plots showing the range of NSE values for the 8 calibration sites, for the calibration period (a) and validation period (b) for the snow threshold Ts = −0.5 ℃ and the rain threshold Tr = 3 ℃. (c) Changes in the average NSE values as a function of Ts and Tr. Here, the median NSE values were averaged to obtain the values shown in this figure. The chosen parameter combination of Ts = −0.5 ℃ and Tr = 3 ℃ are marked by a black dot. (d) Box plot of the NSE values for each of the 8 calibration sites for the chosen thresholds over the entire study period (Adopted from Pelak et al., 2022) Efforts to link remote sensing, hydrological modeling, and in situ observations have also been made to better understand the spatial-temporal behavior of seasonal snow cover. For example, a novel algorithm has been developed to improve the quality of remotely sensed snow datasets by incorporating ground-based meteorological observations, leading to more accurate snow cover simulations (Dong et al., 2016). 6.4 Applications in predictive water resource management The advancements in snowpack measurement and modeling have significant implications for predictive water resource management. Accurate snowpack estimates are essential for forecasting water availability, particularly in regions where snowmelt is a critical water source. For instance, in the Elqui River watershed in Chile, a snowmelt forecast model using remotely-sensed snow cover products has been developed to predict future water availability despite minimal in situ measurements (Sproles et al., 2016).
RkJQdWJsaXNoZXIy MjQ4ODYzMg==