Statistical downscaling with the downscaleR package (v3.1.0): contribution to the VALUE intercomparison experiment

Abstract

Abstract. The increasing demand for high-resolution climate information has attracted growing attention to statistical downscaling (SDS) methods, due in part to their relative advantages and merits as compared to dynamical approaches (based on regional climate model simulations), such as their much lower computational cost and their fitness for purpose for many local-scale applications. As a result, a plethora of SDS methods is nowadays available to climate scientists, which has motivated recent efforts for their comprehensive evaluation, like the VALUE initiative (http://www.value-cost.eu, last access: 29 March 2020). The systematic intercomparison of a large number of SDS techniques undertaken in VALUE, many of them independently developed by different authors and modeling centers in a variety of languages/environments, has shown a compelling need for new tools allowing for their application within an integrated framework. In this regard, downscaleR is an R package for statistical downscaling of climate information which covers the most popular approaches (model output statistics – including the so-called “bias correction” methods – and perfect prognosis) and state-of-the-art techniques. It has been conceived to work primarily with daily data and can be used in the framework of both seasonal forecasting and climate change studies. Its full integration within the climate4R framework (Iturbide et al., 2019) makes possible the development of end-to-end downscaling applications, from data retrieval to model building, validation, and prediction, bringing to climate scientists and practitioners a unique comprehensive framework for SDS model development. In this article the main features of downscaleR are showcased through the replication of some of the results obtained in VALUE, placing an emphasis on the most technically complex stages of perfect-prognosis model calibration (predictor screening, cross-validation, and model selection) that are accomplished through simple commands allowing for extremely flexible model tuning, tailored to the needs of users requiring an easy interface for different levels of experimental complexity. As part of the open-source climate4R framework, downscaleR is freely available and the necessary data and R scripts to fully replicate the experiments included in this paper are also provided as a companion notebook.

Configuration and intercomparison of deep learning neural models for statistical downscaling

“…For temperature, the output is the mean of a Gaussian distribution (one output node for each target grid box) and training is performed by minimizing the mean square error. For precipitation, due to its mixed discrete-continuous nature, the network optimizes the negative log likelihood of a Bernoulligamma distribution following the approach previously intro-duced by Cannon (2008). In particular, the network estimates the parameter p (i.e., probability of rain) of the Bernoulli distribution for rain occurrence and the parameters α (shape) and β (scale) of the gamma rain amount model, as illustrated in the output layer of Fig.…”

Statistical downscaling of daily temperature and precipitation over China using deep learning neural models: Localization and comparison with other methods

“…As an extension of classical linear regression model (Nelder and Wedderburn, 1972), the GLM code used for this study is developed by Bedia et al . (2020) in the downscaleR package. It assumes that the relationship between the expected value E of a dependent variable y (typically following a particular distribution in an exponential family) and the independent predictors x j can be expressed as: where g is the link function and β j are a set of model parameters.…”

“…(2020), Bedia et al . (2020), and James Hiebert for developing CNN model codes (https://github.com/SantanderMetGroup/DeepDownscaling), downscaleR R‐package (https://github.com/SantanderMetGroup/downscaleR), and ClimDown R‐package (https://github.com/pacificclimate/ClimDown) and making them available. The simulations and analyses were conducted on supercomputers maintained by Climate, Environment and Sustainability Center, Nanjing University of Information Science and Technology.…”

“…This research was supported by Meteorological Research Program of Guangxi (2019TG03) and China Scholar Council fellowship to Sun for his visit at the University of Maryland. We thank the Institute of Tibetan Plateau Research for providing high-resolution observation datasets (http://westdc.westgis.ac.cn/data/7a35329c-c53f-4267-aa07-e0037d913a21), Computational and Information Systems Laboratory of National Center for Atmospheric Research for providing ERI reanalysis, Baño-Medina et al (2020), Bedia et al (2020), and James Hiebert for developing CNN model codes (https://github. com/SantanderMetGroup/DeepDownscaling), down-scaleR R-package (https://github.com/SantanderMetGro up/downscaleR), and ClimDown R-package (https:// github.com/pacificclimate/ClimDown) and making them available.…”

Investigation of The Effect of Climate Change on Extreme Precipitation: Tekirdağ Case

“…In order to close the information gap regarding future climate, currently Global Climate Models (GCMs) are the most advanced tools. GCM data is widely used for the climate change studies however these raw GCM data is quite coarse for local scale studies and impact evaluations because of resolution problem, inconsistent physical processes, and regional patterns (Emami & Koch, 2018;Bedia et. al., 2020).…”