Abstract
Downscaling considerably alleviates the drawbacks of regional climate simulation by general circulation models (GCMs). However, little information is available regarding the downscaling using machine learning methods, specifically at hydrological basin scale. This study developed multiple machine learning (ML) downscaling models, based on a Bayesian model average (BMA), to downscale the precipitation simulation of 8 Coupled Model Intercomparison Project Phase 5 (CMIP5) models using model output statistics (MOS) for the years 1961–2005 in the upper Han River basin. A series of statistical metrics, including Pearson’s correlation coefficient (PCC), root mean squared error (RMSE), and relative bias (Rbias), were used for evaluation and comparative analyses. Moreover, the BMA and the best ML downscaling model were used to downscale precipitation in the 21st century under Representative Concentration Pathway 4.5 (RCP4.5) and RCP8.5 scenarios. The results show the following: (1) The performance of the BMA ensemble simulation is clearly better than that of the individual models and the simple mean model ensemble (MME). The PCC reaches 0.74, and the RMSE is reduced by 28%–60% for all the GCMs and 33% compared to the MME. (2) The downscaled models greatly improved station simulation performance. Support vector machine for regression (SVR) was superior to multilayer perceptron (MLP) and random forest (RF). The downscaling results based on the BMA ensemble simulation and SVR models were regarded as the best performing overall (PCC, RMSE, and Rbias were 0.82, 35.07, mm and −5.45%, respectively). (3) Based on BMA and SVR models, the projected precipitations show a weak increasing trend on the whole under RCP4.5 and RCP8.5. Specifically, the average rainfall during the mid- (2040–2069) and late (2070–2099) 21st century increased by 3.23% and 1.02%, respectively, compared to the base year (1971–2000) under RCP4.5, while they increased by 4.25% and 8.30% under RCP8.5. Additionally, the magnitude of changes during winter and spring was higher than that during summer and autumn. Furthermore, future work is recommended to study the improvement of downscaling models and the effect of local climate.
Highlights
General circulation models (GCMs) are widely utilized to study regional meteorological and hydrological responses under climatic changes. e low resolution of GCMs hinders their applications at a basin scale; downscaling techniques are vital to obtain data at a local scale
ACCESS1.0, GFDL-ESM2M, and GISS-E2-R outperformed the other models in this region, as the Pearson’s correlation coefficient (PCC) reaches 0.64–0.68, and the root mean squared error (RMSE) is approximately 57–82 mm. e good PCCs are paired with poor RMSEs and relative bias (Rbias) in some cases
The differences for each station are apparent. e simulation performance results of all the models for stations 5 and 13 are relatively poor. ese results may be due to a local microclimate as the GCMs cannot consider regional simulation
Summary
General circulation models (GCMs) are widely utilized to study regional meteorological and hydrological responses under climatic changes. e low resolution of GCMs hinders their applications at a basin scale; downscaling techniques are vital to obtain data at a local scale. Support vector machine for regression (SVR) models have been successfully used to capture highly nonlinear relationships by applying kernel functions to map the low-dimensional input data to a high-dimensional feature space [10] Another ML method, random forest (RF), has been regarded as a competent and robust algorithm for representing complex relationships because it can implement different types of input variables and operates flexibly. 3. Methods is study aimed to develop an MOS-based BMA_ML ensemble for downscaling models based on the MLP, SVR, and RF methods for the upper Han River basin. E methodology of the present study encompasses the following steps: (i) predictor selection, (ii) GCM ensemble generation, (iii) downscaling using machine learning, and (iv) result evaluation.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
