Assessment of a Spatiotemporal Deep Learning Approach for Soil Moisture Prediction and Filling the Gaps in Between Soil Moisture Observations

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.

The use of remote sensing indices to understand flood‐recharged soil moisture impacts on trees in semi‐arid floodplains: A review

“…However, pre‐flood soil moisture also influences flooding. For instance, ElSaadani et al (2021) noted that soil mature determines the likelihood of flooding, while Wasko and Nathan (2019) concluded that changes in soil moisture and rainfall influence flood trend in Australia. Figure 1 depicts the conceptual comparison of the post‐flood results between precipitation‐induced infiltration soil moisture and precipitation plus deep infiltration FRSM on depth for semi‐arid local area south of the Equator.…”

“…However, preflood soil moisture also influences flooding. For instance,ElSaadani et al (2021) noted that soil mature determines the likelihood of flooding, whileWasko and Nathan (2019) concluded that changes in soil moisture and rainfall influence flood trend in Australia.…”

Applying a Wavelet Transform Technique to Optimize General Fitting Models for SM Analysis: A Case Study in Downscaling over the Qinghai–Tibet Plateau

“…In addition, a large number of empirical or semi-physical relationships have been developed to link coarse-resolution, satellite-based SM retrievals with other land-surface parameters to obtain finer-resolution and more-continuous SM data products. These fitted models take a wide variety of forms, including optical and thermal temperature/vegetation feature space regression [18][19][20], active and passive microwave data fusion [16,[21][22][23], machine learning [24][25][26], deep learning [27,28], geostatistical methods [29][30][31], and data assimilation methods [32][33][34]. ElSaadani et al investigated the applicability of a convolutional long short-term memory network (ConvLSTM) algorithm for predicting SM and filling observational gaps in south Louisiana in the United States [28].…”

“…These fitted models take a wide variety of forms, including optical and thermal temperature/vegetation feature space regression [18][19][20], active and passive microwave data fusion [16,[21][22][23], machine learning [24][25][26], deep learning [27,28], geostatistical methods [29][30][31], and data assimilation methods [32][33][34]. ElSaadani et al investigated the applicability of a convolutional long short-term memory network (ConvLSTM) algorithm for predicting SM and filling observational gaps in south Louisiana in the United States [28]. Prasad et al designed a new multivariate sequential predictive model that utilizes the ensemble empirical mode decomposition (EEMD) algorithm hybridized with extreme learning machines (ELMs) to forecast soil moisture (SM) over weekly horizons [35].…”

Efficient data-driven gap filling of satellite image time series using deep neural networks with partial convolutions

“…Deep learning (DL) and particularly convolutional neural networks (CNNs) have been exceptionally successful for satellite image analysis tasks including object detection and segmentation. In recent years, DL models have been increasingly used also for modeling continuous spatiotemporal phenomena such as soil moisture (ElSaadani et al, 2021) or air temperature (Amato et al, 2020). Developments in De Bézenac et al (2019) demonstrate that such models can reproduce fundamental physical properties of processes such as advection and diffusion in a purely data-driven way.…”