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

Qinghua Miao,Hao Wang,Kuolin Hsu,Baoxiang Pan,Soroosh Sorooshian

doi:10.3390/w11050977

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 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.

Highlights

Precipitation is a primary force in hydrological systems [1]
To take the advantages from both Convolutional Neural Networks (CNNs) and Long Short Term Memory (LSTM), this study develops a deep neural network composed of convolution layers and the LSTM recurrent module to estimate precipitation based on well-resolved atmospheric dynamical fields
The geopotential height might be the most useful one. Considering both the accuracy and complexity of the model, we suggest that the combination of geopotential height and total water vapor might be reasonable

Summary

Introduction

Precipitation is a primary force in hydrological systems [1]. Obtaining accurate and reliable precipitation data at relevant spatial and temporal scales is crucial for efficient water resources management and timely warning of precipitation-related natural hazards, such as flood andWater 2019, 11, 977; doi:10.3390/w11050977 www.mdpi.com/journal/waterWater 2019, 11, 977 drought [2,3]. Obtaining accurate and reliable precipitation data at relevant spatial and temporal scales is crucial for efficient water resources management and timely warning of precipitation-related natural hazards, such as flood and. To sustain a reasonably long lead-time for the above-mentioned applications, it is imperative to employ precipitation prediction techniques. For short-term range up to climate range, numerical weather/climate modeling is perhaps the only reliable tool for predictions. Numerical models have achieved impressive progress in predicting atmospheric dynamics and physics [4]. Dynamics refers to atmospheric state variables (i.e., density, pressure, temperature, and velocity) that are explicitly described by atmospheric primitive equations and resolved by numerical partial differential equation solvers, while physics refers to the unresolved processes that are diagnosed from the resolved variables based on empirical parameterization schemes.

Methods

Results

Conclusion