skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A Two-Stage Deep Neural Network Framework for Precipitation Estimation from Bispectral Satellite Information

Abstract

Compared to ground precipitation measurements, satellite-based precipitation estimation products have the advantage of global coverage and high spatiotemporal resolutions. However, the accuracy of satellite-based precipitation products is still insufficient to serve many weather, climate, and hydrologic applications at high resolutions. In this paper, the authors develop a state-of-the-art deep learning framework for precipitation estimation using bispectral satellite information, infrared (IR), and water vapor (WV) channels. Specifically, a two-stage framework for precipitation estimation from bispectral information is designed, consisting of an initial rain/no-rain (R/NR) binary classification, followed by a second stage estimating the nonzero precipitation amount. In the first stage, the model aims to eliminate the large fraction of NR pixels and to delineate precipitation regions precisely. In the second stage, the model aims to estimate the pointwise precipitation amount accurately while preserving its heavily skewed distribution. Stacked denoising autoencoders (SDAEs), a commonly used deep learning method, are applied in both stages. Performance is evaluated along a number of common performance measures, including both R/NR and real-valued precipitation accuracy, and compared with an operational product, Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks–Cloud Classification System (PERSIANN-CCS). For R/NR binary classification, the proposed two-stage model outperforms PERSIANN-CCS by 32.56% inmore » the critical success index (CSI). For real-valued precipitation estimation, the two-stage model is 23.40% lower in average bias, is 44.52% lower in average mean squared error, and has a 27.21% higher correlation coefficient. Hence, the two-stage deep learning framework has the potential to serve as a more accurate and more reliable satellite-based precipitation estimation product. Finally, the authors provide some future directions for development of satellite-based precipitation estimation products in both incorporating auxiliary information and improving retrieval algorithms.« less

Authors:
 [1];  [1];  [2];  [1];  [1]
  1. Center for Hydrometeorology and Remote Sensing, and Department of Civil and Environmental Engineering, University of California, Irvine, Irvine, California
  2. Department of Computer Science, University of California, Irvine, Irvine, California
Publication Date:
Research Org.:
Univ. of California, Oakland, CA (United States)
Sponsoring Org.:
USDOE Office of International Affairs (IA)
OSTI Identifier:
1420620
Alternate Identifier(s):
OSTI ID: 1541858
Grant/Contract Number:  
IA0000018
Resource Type:
Published Article
Journal Name:
Journal of Hydrometeorology
Additional Journal Information:
Journal Name: Journal of Hydrometeorology Journal Volume: 19 Journal Issue: 2; Journal ID: ISSN 1525-755X
Publisher:
American Meteorological Society
Country of Publication:
United States
Language:
English
Subject:
54 ENVIRONMENTAL SCIENCES; meteorology & atmospheric sciences; precipitation; data mining; remote sensing; neural networks

Citation Formats

Tao, Yumeng, Hsu, Kuolin, Ihler, Alexander, Gao, Xiaogang, and Sorooshian, Soroosh. A Two-Stage Deep Neural Network Framework for Precipitation Estimation from Bispectral Satellite Information. United States: N. p., 2018. Web. doi:10.1175/JHM-D-17-0077.1.
Tao, Yumeng, Hsu, Kuolin, Ihler, Alexander, Gao, Xiaogang, & Sorooshian, Soroosh. A Two-Stage Deep Neural Network Framework for Precipitation Estimation from Bispectral Satellite Information. United States. doi:10.1175/JHM-D-17-0077.1.
Tao, Yumeng, Hsu, Kuolin, Ihler, Alexander, Gao, Xiaogang, and Sorooshian, Soroosh. Thu . "A Two-Stage Deep Neural Network Framework for Precipitation Estimation from Bispectral Satellite Information". United States. doi:10.1175/JHM-D-17-0077.1.
@article{osti_1420620,
title = {A Two-Stage Deep Neural Network Framework for Precipitation Estimation from Bispectral Satellite Information},
author = {Tao, Yumeng and Hsu, Kuolin and Ihler, Alexander and Gao, Xiaogang and Sorooshian, Soroosh},
abstractNote = {Compared to ground precipitation measurements, satellite-based precipitation estimation products have the advantage of global coverage and high spatiotemporal resolutions. However, the accuracy of satellite-based precipitation products is still insufficient to serve many weather, climate, and hydrologic applications at high resolutions. In this paper, the authors develop a state-of-the-art deep learning framework for precipitation estimation using bispectral satellite information, infrared (IR), and water vapor (WV) channels. Specifically, a two-stage framework for precipitation estimation from bispectral information is designed, consisting of an initial rain/no-rain (R/NR) binary classification, followed by a second stage estimating the nonzero precipitation amount. In the first stage, the model aims to eliminate the large fraction of NR pixels and to delineate precipitation regions precisely. In the second stage, the model aims to estimate the pointwise precipitation amount accurately while preserving its heavily skewed distribution. Stacked denoising autoencoders (SDAEs), a commonly used deep learning method, are applied in both stages. Performance is evaluated along a number of common performance measures, including both R/NR and real-valued precipitation accuracy, and compared with an operational product, Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks–Cloud Classification System (PERSIANN-CCS). For R/NR binary classification, the proposed two-stage model outperforms PERSIANN-CCS by 32.56% in the critical success index (CSI). For real-valued precipitation estimation, the two-stage model is 23.40% lower in average bias, is 44.52% lower in average mean squared error, and has a 27.21% higher correlation coefficient. Hence, the two-stage deep learning framework has the potential to serve as a more accurate and more reliable satellite-based precipitation estimation product. Finally, the authors provide some future directions for development of satellite-based precipitation estimation products in both incorporating auxiliary information and improving retrieval algorithms.},
doi = {10.1175/JHM-D-17-0077.1},
journal = {Journal of Hydrometeorology},
number = 2,
volume = 19,
place = {United States},
year = {2018},
month = {2}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record
DOI: 10.1175/JHM-D-17-0077.1

Citation Metrics:
Cited by: 2 works
Citation information provided by
Web of Science

Save / Share: