Abstract
Climate downscaling is a way to provide finer resolution data at local scales, which has been widely used in meteorological research. The two main approaches for climate downscaling are dynamical and statistical. The traditional dynamical downscaling methods are quite time- and resource-consuming based on general circulation models (GCMs). Recently, more and more researchers construct a statistical deep learning model for climate downscaling motivated by the single-image superresolution (SISR) process in computer vision (CV). This is an approach that uses historical climate observations to learn a low-resolution to high-resolution mapping and produces great enhancements in terms of efficiency and effectiveness. Therefore, it has provided an appreciable new insight and successful downscaling solution to multiple climate phenomena. However, most existing models only make a simple analogy between climate downscaling and SISR and ignore the underlying dynamical mechanisms, which leads to the overaveraged downscaling results lacking crucial physical details. In this paper, we incorporate the a priori meteorological knowledge into a deep learning formalization for climate downscaling. More specifically, we consider the multiscale spatial correlations and the chaos in multiple climate events. Depending on two characteristics, we build up a two-stage deep learning model containing a stepwise reconstruction process and ensemble inference, which is named climate downscaling network (CDN). It can extract more local/remote spatial dependencies and provide more comprehensive captures of extreme conditions. We evaluate our model based on two datasets: climate science dataset (CSD) and benchmark image dataset (BID). The results demonstrate that our model shows the effectiveness and superiority in downscaling daily precipitation data from 2.5 degrees to 0.5 degrees over Asia and Europe. In addition, our model exhibits better performance than the other traditional approaches and state-of-the-art deep learning models.
Highlights
Climate variations are influencing society’s well-being all over the world, such as global warming, extreme storm, precipitation, and sea-level rising [1,2,3]
General circulation models (GCMs) are the traditional tools for simulating and investigating the various climate events [6, 7]. ey are usually developed on the basis of complicated physical dynamical principles and executed on large-scale supercomputers. ese GCMs usually take a large number of physical variables into account and model the entire climate evolution processes into specific differential equations
2 Benchmark image dataset (BID): this dataset is composed of several different datasets. e training and validation set is from DIV2K [36], containing 800 images for training and 100 images for validation, respectively. e testing set is from 5 datasets: SET5 [37], SET14 [38], BSDS100 [39], URBAN100 [40], and MANGA109 [41]
Summary
Climate variations are influencing society’s well-being all over the world, such as global warming, extreme storm, precipitation, and sea-level rising [1,2,3]. General circulation models (GCMs) are the traditional tools for simulating and investigating the various climate events [6, 7]. Ese GCMs usually take a large number of physical variables into account and model the entire climate evolution processes into specific differential equations. Ese variables are treated as initial fields and have specific spatial resolutions, respectively, so that the meteorological researchers can investigate and forecast the potential trends that rely on these physical patterns with various resolutions under model integration. When simulating and studying the meteorological evolutions in small-scale areas, one may encounter the issues of too coarse spatial resolutions in GCMs, which limit revealing critical detailed physical processes, such as precipitations in key river basins and extreme typhoons. When simulating and studying the meteorological evolutions in small-scale areas, one may encounter the issues of too coarse spatial resolutions in GCMs, which limit revealing critical detailed physical processes, such as precipitations in key river basins and extreme typhoons. is severely hinders the assessment of the impact of climate events on these relevant key regions [8].
Sign-in/Register to view highlights & summary 
Published Version (
Free)
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 call