With many operational satellite rainfall products being available for long periods, it is now possible to examine whether these products can reproduce climatologically known rainfall characteristics over large river basins that suffer from poor surface monitoring resources. Such assessment is a prerequisite for any further hydrologic applications that rely on these products. The current study evaluates two satellite rainfall products, the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis research-grade product (TMPA-3B42) and the Climate Prediction Center (CPC) morphing technique (CMORPH), over the domain of the Nile basin in eastern Africa. The large latitudinal extent of the basin, its complex topography, and its diverse land use result in widely contrasting regimes and distributions of annual and seasonal rainfall. The results suggest that the two products are fairly successful in reproducing some of the region-specific rainfall patterns across climatologically different parts of the basin. However, significant overestimation and underestimation by CMORPH and TMPA-3B42, respectively, are clearly evident over the majority of the basin and can exceed 100% of the mean annual rainfall. The biases are also evident in the seasonal rainfall cycle. The bias shows a complex dependency, in terms of magnitude and sign, on topography and latitude, especially in the central parts of the basin and over the Ethiopian Highlands region. The performance of both products is better over the equatorial region of the basin. The significant underestimation by the gauge-adjusted TMPA-3B42 product relative to CMORPH is attributed to the sparsity in operational gauges, which can adversely affect bias adjustment procedures in the TMPA algorithm. Current and future algorithmic developments are expected to bring much-needed improvements for satellite rainfall products to see full operational utility in these regions of the world.

No abstract

Soil moisture (SM) plays a significant role in determining the probability of flooding in a given area. Currently, SM is most commonly modeled using physically-based numerical hydrologic models. Modeling the natural processes that take place in the soil is difficult and requires assumptions. Besides, hydrologic model runtime is highly impacted by the extent and resolution of the study domain. In this study, we propose a data-driven modeling approach using Deep Learning (DL) models. There are different types of DL algorithms that serve different purposes. For example, the Convolutional Neural Network (CNN) algorithm is well suited for capturing and learning spatial patterns, while the Long Short-Term Memory (LSTM) algorithm is designed to utilize time-series information and to learn from past observations. A DL algorithm that combines the capabilities of CNN and LSTM called ConvLSTM was recently developed. In this study, we investigate the applicability of the ConvLSTM algorithm in predicting SM in a study area located in south Louisiana in the United States. This study reveals that ConvLSTM significantly outperformed CNN in predicting SM. We tested the performance of ConvLSTM based models by using a combination of different sets of predictors and different LSTM sequence lengths. The study results show that ConvLSTM models can predict SM with a mean areal Root Mean Squared Error (RMSE) of 2.5% and mean areal correlation coefficients of 0.9 for our study area. ConvLSTM models can also provide predictions between discrete SM observations, making them potentially useful for applications such as filling observational gaps between satellite overpasses.

This study describes the generation and testing of a reference rainfall product created from field campaign datasets collected during the NASA Global Precipitation Measurement (GPM) mission Ground Validation Iowa Flood Studies (IFloodS) experiment. The study evaluates ground-based radar rainfall (RR) products acquired during IFloodS in the context of building the reference rainfall product. The purpose of IFloodS was not only to attain a high-quality ground-based reference for the validation of satellite rainfall estimates but also to enhance understanding of flood-related rainfall processes and the predictability of flood forecasting. We assessed the six RR estimates (IFC, Q2, CSU-DP, NWS-DP, Stage IV, and Q2-Corrected) using data from rain gauge and disdrometer networks that were located in the broader field campaign area of central and northeastern Iowa. We performed the analyses with respect to time scales ranging from 1 h to the entire campaign period in order to compare the capabilities of each RR product and to characterize the error structure at scales that are frequently used in hydrologic applications. The evaluation results show that the Stage IV estimates perform superior to other estimates, demonstrating the need for gauge-based bias corrections of radar-only products. This correction should account for each product’s algorithm-dependent error structure that can be used to build unbiased rainfall products for the campaign reference. We characterized the statistical error structures (e.g., systematic and random components) of each RR estimate and used them for the generation of a campaign reference rainfall product. To assess the hydrologic utility of the reference product, we performed hydrologic simulations driven by the reference product over the Turkey River basin. The comparison of hydrologic simulation results demonstrates that the campaign reference product performs better than Stage IV in streamflow generation.

Radar-rainfall products provide valuable information for hydro-ecological modeling and ecosystem applications, especially over coastal regions that lack adequate in-situ rainfall observations. This study evaluates two radar-based rainfall products, the Multi-Sensor Stage IV and the Multi-Radar Multi-Sensor (MRMS), over the Louisiana coastal region in the United States. Surface reference rainfall observations from two independent rain gage networks were used in the analysis. The evaluation included distribution-based comparisons between radar and gage observations at different time scales (hourly to monthly), bias decomposition to quantify the contribution of different error sources, and conditional evaluation of systematic and random components of the estimation errors. Both products report large levels of random errors at the hourly scale; however, the performance of the radar-rainfall products improves significantly with the increase in time scales. After decomposing the total bias, the results show that the largest contributor to the overall bias in radar-rainfall products is false rainfall detection, followed by missed rainfall. The results also reveal that the Stage IV product experienced a significant improvement over the area in the past few years (post 2015) compared to earlier periods. The results have implications for ongoing and future coastal ecosystem modeling and planning studies.