DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Improving Monsoon Precipitation Prediction Using Combined Convolutional and Long Short Term Memory Neural Network

Abstract

Precipitation downscaling is widely employed for enhancing the resolution and accuracy of precipitation products from general circulation models (GCMs). In this study, we propose a novel statistical downscaling method to foster GCMs’ precipitation prediction resolution and accuracy for the monsoon region. We develop a deep neural network composed of a convolution and Long Short Term Memory (LSTM) recurrent module to estimate precipitation based on well-resolved atmospheric dynamical fields. The proposed model is compared against the GCM precipitation product and classical downscaling methods in the Xiangjiang River Basin in South China. Results show considerable improvement compared to the European Centre for Medium-Range Weather Forecasts (ECMWF)-Interim reanalysis precipitation. Also, the model outperforms benchmark downscaling approaches, including (1) quantile mapping, (2) the support vector machine, and (3) the convolutional neural network. To test the robustness of the model and its applicability in practical forecasting, we apply the trained network for precipitation prediction forced by retrospective forecasts from the ECMWF model. Compared to the ECMWF precipitation forecast, our model makes better use of the resolved dynamical field for more accurate precipitation prediction at lead times from 1 day up to 2 weeks. This superiority decreases with the forecast lead time, as the GCM’s skillmore » in predicting atmospheric dynamics is diminished by the chaotic effect. Finally, we build a distributed hydrological model and force it with different sources of precipitation inputs. Hydrological simulation forced with the neural network precipitation estimation shows significant advantage over simulation forced with the original ERA-Interim precipitation (with NSE value increases from 0.06 to 0.64), and the performance is only slightly worse than the observed precipitation forced simulation (NSE = 0.82). This further proves the value of the proposed downscaling method, and suggests its potential for hydrological forecasts.« less

Authors:
 [1];  [2];  [3];  [2];  [2]
  1. Tsinghua Univ., Beijing (China)
  2. Univ. of California, Irvine, CA (United States)
  3. Tsinghua Univ., Beijing (China); China Inst. of Water Resources and Hydropower Research, Beijing (China)
Publication Date:
Research Org.:
Univ. of California, Oakland, CA (United States)
Sponsoring Org.:
USDOE Office of International Affairs (IA); National Natural Science Foundation of China (NSFC); China Scholarship Council
OSTI Identifier:
1613794
Grant/Contract Number:  
IA0000018; 41661144031
Resource Type:
Accepted Manuscript
Journal Name:
Water (Basel)
Additional Journal Information:
Journal Name: Water (Basel); Journal Volume: 11; Journal Issue: 5; Journal ID: ISSN 2073-4441
Publisher:
MDPI
Country of Publication:
United States
Language:
English
Subject:
54 ENVIRONMENTAL SCIENCES; Water Resources; precipitation downscaling; convolutional neural networks; long short term memory networks; hydrological simulation

Citation Formats

Miao, Qinghua, Pan, Baoxiang, Wang, Hao, Hsu, Kuolin, and Sorooshian, Soroosh. Improving Monsoon Precipitation Prediction Using Combined Convolutional and Long Short Term Memory Neural Network. United States: N. p., 2019. Web. doi:10.3390/w11050977.
Miao, Qinghua, Pan, Baoxiang, Wang, Hao, Hsu, Kuolin, & Sorooshian, Soroosh. Improving Monsoon Precipitation Prediction Using Combined Convolutional and Long Short Term Memory Neural Network. United States. https://doi.org/10.3390/w11050977
Miao, Qinghua, Pan, Baoxiang, Wang, Hao, Hsu, Kuolin, and Sorooshian, Soroosh. Thu . "Improving Monsoon Precipitation Prediction Using Combined Convolutional and Long Short Term Memory Neural Network". United States. https://doi.org/10.3390/w11050977. https://www.osti.gov/servlets/purl/1613794.
@article{osti_1613794,
title = {Improving Monsoon Precipitation Prediction Using Combined Convolutional and Long Short Term Memory Neural Network},
author = {Miao, Qinghua and Pan, Baoxiang and Wang, Hao and Hsu, Kuolin and Sorooshian, Soroosh},
abstractNote = {Precipitation downscaling is widely employed for enhancing the resolution and accuracy of precipitation products from general circulation models (GCMs). In this study, we propose a novel statistical downscaling method to foster GCMs’ precipitation prediction resolution and accuracy for the monsoon region. We develop a deep neural network composed of a convolution and Long Short Term Memory (LSTM) recurrent module to estimate precipitation based on well-resolved atmospheric dynamical fields. The proposed model is compared against the GCM precipitation product and classical downscaling methods in the Xiangjiang River Basin in South China. Results show considerable improvement compared to the European Centre for Medium-Range Weather Forecasts (ECMWF)-Interim reanalysis precipitation. Also, the model outperforms benchmark downscaling approaches, including (1) quantile mapping, (2) the support vector machine, and (3) the convolutional neural network. To test the robustness of the model and its applicability in practical forecasting, we apply the trained network for precipitation prediction forced by retrospective forecasts from the ECMWF model. Compared to the ECMWF precipitation forecast, our model makes better use of the resolved dynamical field for more accurate precipitation prediction at lead times from 1 day up to 2 weeks. This superiority decreases with the forecast lead time, as the GCM’s skill in predicting atmospheric dynamics is diminished by the chaotic effect. Finally, we build a distributed hydrological model and force it with different sources of precipitation inputs. Hydrological simulation forced with the neural network precipitation estimation shows significant advantage over simulation forced with the original ERA-Interim precipitation (with NSE value increases from 0.06 to 0.64), and the performance is only slightly worse than the observed precipitation forced simulation (NSE = 0.82). This further proves the value of the proposed downscaling method, and suggests its potential for hydrological forecasts.},
doi = {10.3390/w11050977},
journal = {Water (Basel)},
number = 5,
volume = 11,
place = {United States},
year = {Thu May 09 00:00:00 EDT 2019},
month = {Thu May 09 00:00:00 EDT 2019}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 57 works
Citation information provided by
Web of Science

Save / Share:

Works referenced in this record:

Application of a distributed hydrological model and weather radar observations for flood management in the upper Tone River of Japan
journal, January 2004

  • Yang, Dawen; Koike, Toshio; Tanizawa, Hiroshi
  • Hydrological Processes, Vol. 18, Issue 16
  • DOI: 10.1002/hyp.5752

Ensemble Simulations of Asian–Australian Monsoon Variability by 11 AGCMs*
journal, February 2004


Information Analysis of Catchment Hydrologic Patterns across Temporal Scales
journal, January 2016


The Subseasonal to Seasonal (S2S) Prediction Project Database
journal, January 2017

  • Vitart, F.; Ardilouze, C.; Bonet, A.
  • Bulletin of the American Meteorological Society, Vol. 98, Issue 1
  • DOI: 10.1175/BAMS-D-16-0017.1

Statistical downscaling of climate scenarios over Scandinavia
journal, January 2005

  • Hanssen-Bauer, I.; Achberger, C.; Benestad, Re
  • Climate Research, Vol. 29
  • DOI: 10.3354/cr029255

Deep learning
journal, May 2015

  • LeCun, Yann; Bengio, Yoshua; Hinton, Geoffrey
  • Nature, Vol. 521, Issue 7553
  • DOI: 10.1038/nature14539

Validation and comparison of a new gauge-based precipitation analysis over mainland China: NEW GAUGE-BASED PRECIPITATION OVER CHINA
journal, April 2015

  • Shen, Yan; Xiong, Anyuan
  • International Journal of Climatology, Vol. 36, Issue 1
  • DOI: 10.1002/joc.4341

Projected future changes in synoptic systems influencing southwest Western Australia
journal, February 2006


A distributed scheme developed for eco-hydrological modeling in the upper Heihe River
journal, December 2014


Downscaling temperature and precipitation: a comparison of regression-based methods and artificial neural networks
journal, January 2001

  • Schoof, J. T.; Pryor, S. C.
  • International Journal of Climatology, Vol. 21, Issue 7
  • DOI: 10.1002/joc.655

A Statistical Downscaling Model for Southern Australia Winter Rainfall
journal, March 2009


Validation of present-day regional climate simulations over Europe: LAM simulations with observed boundary conditions
journal, August 1997

  • Christensen, J. H.; Machenhauer, B.; Jones, R. G.
  • Climate Dynamics, Vol. 13, Issue 7-8
  • DOI: 10.1007/s003820050178

Is Precipitation a Good Metric for Model Performance?
journal, February 2019

  • Tapiador, Francisco J.; Roca, Rémy; Del Genio, Anthony
  • Bulletin of the American Meteorological Society, Vol. 100, Issue 2
  • DOI: 10.1175/BAMS-D-17-0218.1

Role of predictors in downscaling surface temperature to river basin in India for IPCC SRES scenarios using support vector machine
journal, March 2009

  • Anandhi, Aavudai; Srinivas, V. V.; Kumar, D. Nagesh
  • International Journal of Climatology, Vol. 29, Issue 4
  • DOI: 10.1002/joc.1719

Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition
journal, June 1965


Development of a Geomorphology-Based Hydrological Model for Large Catchments
journal, January 1998

  • Yang, Dawen; Herath, Srikantha; Musiake, Katumi
  • PROCEEDINGS OF HYDRAULIC ENGINEERING, Vol. 42
  • DOI: 10.2208/prohe.42.169

A Comparison of Statistical and Model-Based Downscaling Techniques for Estimating Local Climate Variations
journal, April 1998


Flood Prediction Using Machine Learning Models: Literature Review
journal, October 2018

  • Mosavi, Amir; Ozturk, Pinar; Chau, Kwok-wing
  • Water, Vol. 10, Issue 11
  • DOI: 10.3390/w10111536

Sensitivity of flood events to global climate change
journal, April 1997


Long Short-Term Memory
journal, November 1997


The ERA-Interim reanalysis: configuration and performance of the data assimilation system
journal, April 2011

  • Dee, D. P.; Uppala, S. M.; Simmons, A. J.
  • Quarterly Journal of the Royal Meteorological Society, Vol. 137, Issue 656
  • DOI: 10.1002/qj.828

A hillslope-based hydrological model using catchment area and width functions
journal, February 2002

  • Yang, Dawen; Herath, Srikantha; Musiake, Katumi
  • Hydrological Sciences Journal, Vol. 47, Issue 1
  • DOI: 10.1080/02626660209492907

Development of a China Dataset of Soil Hydraulic Parameters Using Pedotransfer Functions for Land Surface Modeling
journal, June 2013

  • Dai, Yongjiu; Shangguan, Wei; Duan, Qingyun
  • Journal of Hydrometeorology, Vol. 14, Issue 3
  • DOI: 10.1175/JHM-D-12-0149.1

Novel SVM-based technique to improve rainfall estimation over the Mediterranean region (north of Algeria) using the multispectral MSG SEVIRI imagery
journal, March 2017


Downscaling of global climate models for flood frequency analysis: where are we now?
journal, January 2002

  • Prudhomme, Christel; Reynard, Nick; Crooks, Sue
  • Hydrological Processes, Vol. 16, Issue 6
  • DOI: 10.1002/hyp.1054

Novel methods for inferring future changes in extreme rainfall over Northern Europe
journal, September 2007


SVM-PGSL coupled approach for statistical downscaling to predict rainfall from GCM output
journal, January 2010


On complex extremes: flood hazards and combined high spring-time precipitation and temperature in Norway
journal, June 2007


The quiet revolution of numerical weather prediction
journal, September 2015

  • Bauer, Peter; Thorpe, Alan; Brunet, Gilbert
  • Nature, Vol. 525, Issue 7567
  • DOI: 10.1038/nature14956

Downscaling of precipitation for climate change scenarios: A support vector machine approach
journal, November 2006


A comparative assessment of decision trees algorithms for flash flood susceptibility modeling at Haraz watershed, northern Iran
journal, June 2018


Projecting changes in future heavy rainfall events for Oahu, Hawaii: A statistical downscaling approach
journal, January 2011

  • Norton, Chase W.; Chu, Pao-Shin; Schroeder, Thomas A.
  • Journal of Geophysical Research, Vol. 116, Issue D17
  • DOI: 10.1029/2011JD015641

Spatial resolution sensitivity of catchment geomorphologic properties and the effect on hydrological simulation
journal, January 2001

  • Yang, Dawen; Herath, Srikantha; Musiake, Katumi
  • Hydrological Processes, Vol. 15, Issue 11
  • DOI: 10.1002/hyp.280

Dreary state of precipitation in global models: MODEL AND OBSERVED PRECIPITATION
journal, December 2010

  • Stephens, Graeme L.; L'Ecuyer, Tristan; Forbes, Richard
  • Journal of Geophysical Research: Atmospheres, Vol. 115, Issue D24
  • DOI: 10.1029/2010JD014532

Statistical and dynamical downscaling of the Seine basin climate for hydro-meteorological studies
journal, January 2007

  • Boé, J.; Terray, L.; Habets, F.
  • International Journal of Climatology, Vol. 27, Issue 12
  • DOI: 10.1002/joc.1602

Downscaling precipitation to river basin in India for IPCC SRES scenarios using support vector machine
journal, March 2008

  • Anandhi, Aavudai; Srinivas, V. V.; Nanjundiah, Ravi S.
  • International Journal of Climatology, Vol. 28, Issue 3
  • DOI: 10.1002/joc.1529

An ensemble prediction of flood susceptibility using multivariate discriminant analysis, classification and regression trees, and support vector machines
journal, February 2019


ImageNet classification with deep convolutional neural networks
journal, May 2017

  • Krizhevsky, Alex; Sutskever, Ilya; Hinton, Geoffrey E.
  • Communications of the ACM, Vol. 60, Issue 6
  • DOI: 10.1145/3065386

Improving Precipitation Estimation Using Convolutional Neural Network
journal, March 2019

  • Pan, Baoxiang; Hsu, Kuolin; AghaKouchak, Amir
  • Water Resources Research, Vol. 55, Issue 3
  • DOI: 10.1029/2018WR024090

Precipitation predictors for downscaling: observed and general circulation model relationships
journal, May 2000


Long lead monsoon rainfall prediction for meteorological sub-divisions of India using deterministic artificial neural network model
journal, September 2008


Intercomparison of the climatological variations of Asian summer monsoon precipitation simulated by 10 GCMs
journal, August 2002


From GCMs to river flow: a review of downscaling methods and hydrologic modelling approaches
journal, June 1999


From GCMs to river flow: a review of downscaling methods and hydrologic modelling approaches
journal, June 1999


Improving Precipitation Estimation Using Convolutional Neural Network
text, January 2018


Deep Learning
text, January 2018


Convolutional LSTM Networks for Subcellular Localization of Proteins
text, January 2015


Flood Prediction Using Machine Learning Models: Literature Review
text, January 2019


Works referencing / citing this record:

Hydrological Early Warning System Based on a Deep Learning Runoff Model Coupled with a Meteorological Forecast
journal, August 2019

  • de la Fuente, Alberto; Meruane, Viviana; Meruane, Carolina
  • Water, Vol. 11, Issue 9
  • DOI: 10.3390/w11091808