Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data

Root zone soil moisture (RZSM) is an important hydrological variable for agricultural planning and water resources management. The Soil Moisture Active Passive Level 4 (SMAP L4) data demonstrates great value in RZSM estimation. Accurate sub-seasonal RZSM prediction based on SMAP L4 holds great signi...

Full description

Saved in:
Bibliographic Details
Main Authors: Lei Xu, Xihao Zhang, Xi Zhang, Tingtao Wu, Hongchu Yu, Wenying Du, Zeqiang Chen, Nengcheng Chen
Format: Article
Language:English
Published: Elsevier 2025-05-01
Series:International Journal of Applied Earth Observations and Geoinformation
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1569843225001797
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849321583887253504
author Lei Xu
Xihao Zhang
Xi Zhang
Tingtao Wu
Hongchu Yu
Wenying Du
Zeqiang Chen
Nengcheng Chen
author_facet Lei Xu
Xihao Zhang
Xi Zhang
Tingtao Wu
Hongchu Yu
Wenying Du
Zeqiang Chen
Nengcheng Chen
author_sort Lei Xu
collection DOAJ
description Root zone soil moisture (RZSM) is an important hydrological variable for agricultural planning and water resources management. The Soil Moisture Active Passive Level 4 (SMAP L4) data demonstrates great value in RZSM estimation. Accurate sub-seasonal RZSM prediction based on SMAP L4 holds great significance for agricultural management and drought assessment. Current deep learning-based RZSM prediction models tend to accumulate error in long-term forecasting and the limited SMAP RZSM samples may result in insufficient model generalization. To address these issues, this study proposes a multi-head self-attention-based autoregressive transfer learning model based on long short-term memory (MAATL) model for sub-seasonal RZSM prediction. The proposed MAATL model is evaluated over the Continental United States (CONUS) for 1- to 60-day RZSM prediction and compared with some ablation and long short-term memory (LSTM) models. The results showed that compared with LSTM, the skills of the MAATL model were significantly improved, with an average correlation coefficient increase of 18.26% and a root mean square error (RMSE) reduction of 42.55%. Furthermore, 118 in-situ soil moisture stations are used for predictive validation and the proposed MAATL model demonstrates higher accuracy compared to the Global Forecast System (GFS) and the LSTM model, with an average correlation skill improvement of 16.02% and 15.08% for MAATL over GFS and LSTM, respectively. These findings indicate superior performance for the proposed MAATL model in sub-seasonal RZSM prediction, which has great potential for agricultural drought preparations.
format Article
id doaj-art-ca435fb0c5274ad29a02f76c381fa0df
institution Kabale University
issn 1569-8432
language English
publishDate 2025-05-01
publisher Elsevier
record_format Article
series International Journal of Applied Earth Observations and Geoinformation
spelling doaj-art-ca435fb0c5274ad29a02f76c381fa0df2025-08-20T03:49:42ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322025-05-0113910453210.1016/j.jag.2025.104532Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP dataLei Xu0Xihao Zhang1Xi Zhang2Tingtao Wu3Hongchu Yu4Wenying Du5Zeqiang Chen6Nengcheng Chen7National Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, ChinaNational Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, ChinaNational Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, ChinaNational Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, ChinaSchool of Navigation, Wuhan University of Technology, Wuhan 430063, ChinaNational Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, ChinaNational Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, China; Corresponding authors.National Engineering Research Center for Geographic Information System, China University of Geosciences (Wuhan), Wuhan 430074, ChinaRoot zone soil moisture (RZSM) is an important hydrological variable for agricultural planning and water resources management. The Soil Moisture Active Passive Level 4 (SMAP L4) data demonstrates great value in RZSM estimation. Accurate sub-seasonal RZSM prediction based on SMAP L4 holds great significance for agricultural management and drought assessment. Current deep learning-based RZSM prediction models tend to accumulate error in long-term forecasting and the limited SMAP RZSM samples may result in insufficient model generalization. To address these issues, this study proposes a multi-head self-attention-based autoregressive transfer learning model based on long short-term memory (MAATL) model for sub-seasonal RZSM prediction. The proposed MAATL model is evaluated over the Continental United States (CONUS) for 1- to 60-day RZSM prediction and compared with some ablation and long short-term memory (LSTM) models. The results showed that compared with LSTM, the skills of the MAATL model were significantly improved, with an average correlation coefficient increase of 18.26% and a root mean square error (RMSE) reduction of 42.55%. Furthermore, 118 in-situ soil moisture stations are used for predictive validation and the proposed MAATL model demonstrates higher accuracy compared to the Global Forecast System (GFS) and the LSTM model, with an average correlation skill improvement of 16.02% and 15.08% for MAATL over GFS and LSTM, respectively. These findings indicate superior performance for the proposed MAATL model in sub-seasonal RZSM prediction, which has great potential for agricultural drought preparations.http://www.sciencedirect.com/science/article/pii/S1569843225001797RZSMSMAPMulti-head self-attentionAutoregressiveTransfer learning
spellingShingle Lei Xu
Xihao Zhang
Xi Zhang
Tingtao Wu
Hongchu Yu
Wenying Du
Zeqiang Chen
Nengcheng Chen
Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data
International Journal of Applied Earth Observations and Geoinformation
RZSM
SMAP
Multi-head self-attention
Autoregressive
Transfer learning
title Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data
title_full Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data
title_fullStr Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data
title_full_unstemmed Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data
title_short Accurate sub-seasonal root-zone soil moisture prediction using attention-based autoregressive transfer learning and SMAP data
title_sort accurate sub seasonal root zone soil moisture prediction using attention based autoregressive transfer learning and smap data
topic RZSM
SMAP
Multi-head self-attention
Autoregressive
Transfer learning
url http://www.sciencedirect.com/science/article/pii/S1569843225001797
work_keys_str_mv AT leixu accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT xihaozhang accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT xizhang accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT tingtaowu accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT hongchuyu accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT wenyingdu accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT zeqiangchen accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata
AT nengchengchen accuratesubseasonalrootzonesoilmoisturepredictionusingattentionbasedautoregressivetransferlearningandsmapdata