Publications

Predicting water cycling in agricultural watersheds is challenging due to factors like farming practices. This study looks at using remote sensing evapotranspiration (ET) data and crop yield information to improve the accuracy of the Soil and Water Assessment Tool (SWAT) model. By adding more constraints to the model, such as crop yield, the number of acceptable parameter sets was reduced, and the model's performance improved. The results suggest that using crop yield data as an additional constraint can help reduce uncertainty and increase the accuracy of ET predictions in agricultural watersheds.

Soil moisture is important for understanding the global water cycle, but current satellite measurements are not continuous in time or space. This study combines data from NASA's Cyclone Global Navigation Satellite System (CYGNSS) and the Soil Moisture Active Passive (SMAP) to improve soil moisture estimates in a land surface model (LSM). The results show a 61.3% improvement in LSM soil moisture accuracy when combining the two satellite systems. However, using satellite data in areas with dense vegetation can lead to less accurate results. This research is the first to use global GNSS-based soil moisture observations in LSMs, which can help fill gaps in soil moisture measurements and improve our understanding of the water cycle.

As the intensity and frequency of storm events are projected to increase due to climate change, local agencies urgently need a timely and reliable framework for flood forecasting, downscale from watershed to street level in urban areas. Integrated with property data with various hydrometeorological data, the flood prediction model can also provide further insight into environmental justice, which will aid households and government agencies’ decision-making. This study uses deep learning (DL) methods and radar-based rainfall data to predict the inundated areas and analyze the property quickly and demographic data concerning stream proximity to provide a way to quantify socioeconomic impacts. We expect that our DL-based models will improve the accuracy of forecasting floods and provide a better picture of which communities bear the worst burdens of flooding, and encourage city officials to address the underlying causes of flood risk.

This research focuses on using satellite and modeled products to monitor soil moisture (SM) content and predict natural disasters such as droughts, floods, wildfires, landslides, and dust outbreaks. The study validates three SMAP SM products with in-situ data using conventional and triple collocation analysis (TCA) statistics and merges them with a Noah-Multiparameterization version-3.6 (NoahMP36) land surface model (LSM). An exponential filter and a cumulative density function (CDF) are used to evaluate the SM products. The study found that CDF-matched 9-, 3-, and 1-km SMAP SM data showed reliable performance with R and ubRMSD values of 0.658, 0.626, and 0.570 and 0.049, 0.053, and 0.055 m3/m3, respectively. Combining SMAP and NoahMP36 greatly improved R-values to 0.825, 0.804, and 0.795, and ubRMSDs to 0.034, 0.036, and 0.037 m3/m3, respectively. These findings suggest that SMAP/Sentinel data can improve regional-scale SM estimates and LSMs with improved accuracy.

Hydrologic models have some predictive uncertainty when used for real-world applications. This study looks at using remotely sensed evapotranspiration (RS-ET) data to evaluate improvements in the Soil and Water Assessment Tool (SWAT) model. By comparing the original SWAT model and an improved version (RSWAT), researchers found that both models performed similarly for daily streamflow and evapotranspiration at the watershed level. However, at the subwatershed level, RSWAT showed better results for daily evapotranspiration. This study shows that using RS-ET data can help increase the accuracy of model predictions and highlights the importance of remote sensing data in hydrologic modeling.

This study compares the accuracy and error characteristics of surface soil moisture (SSM) estimates obtained from various satellite and model-based data products over vegetated and irrigated regions. The study employed triple collocation analysis (TCA) and conventional error metrics to evaluate the accuracy of six different products: Advanced Scatterometer (ASCAT), Soil Moisture and Ocean Salinity (SMOS), Advanced Microwave Scanning Radiometer 2 (AMSR2), Soil Moisture Active Passive (SMAP), European Centre for Medium-Range Weather Forecasts Reanalysis 5 (ERA5), and Global Land Data Assimilation System (GLDAS). The results show that satellite-based SSM estimates from ASCAT, SMAP, and SMOS had fewer errors than ERA5 and GLDAS SSM products over vegetated areas, and over irrigated areas, ASCAT, SMOS, and SMAP outperformed other SSM products. The study also found that the limitations in satellite and model-based SSM data can be overcome by the synergistic use of satellite and model-based SSM products. The study suggests that the probability of obtaining SSM with a stronger signal than noise can be close to 100% when four satellite and model data sets are used selectively.