Assessing Water Balance Closure Using Multiple Data Assimilation and Remote Sensing-Based Datasets for Canada

Abstract Obtaining reliable water balance estimates remains a major challenge in Canada for large regions with scarce situ measurements. Various remote sensing products can be used to complement observation-based datasets and provide an estimate of the at river basin or regional scales. This study provides assessment using combinations various data assimilation-based quantifies non-closure errors basins across Canada, ranging from 90,900 1,679,100 km 2 , period 2002 2015. A equation combines following monthly closure: multiple sources each budget component, including two precipitation - global product WATCH Forcing Data ERA-Interim (WFDEI), Canadian Precipitation Analysis (CaPA); evapotranspiration MODIS, Global Land-surface Evaporation: Amsterdam Methodology (GLEAM); one source storage GRACE three different centers; observed discharge hydrometric stations (HYDAT). The error is attributed constrained Kalman filter. Results show that combination CaPA, GLEAM, JPL mascon tended outperform other basins. Overall, attributions precipitation, evapotranspiration, change, runoff were 36.7, 33.2, 17.8, 12.2 percent, which corresponded 8.1, 7.9, 4.2, 1.4 mm month -1 respectively. In particular, dominated Western whereas contributed most Mackenzie River basin.