Climatic Research Unit

Statistical downscaling tools

This page provides information about statistical downscaling tools including a link to the ENSEMBLES downscaling portal.

Statistical downscaling tools

Regional scenarios constructed to assess climate change impacts require finer scales than those provided by global climate models. This transformation is done by dynamical or statistical downscaling. Statistical downscaling involves the application of relationships identified in the observed climate, between the large and smaller-scale, to climate model output. It assumes that the relationships between predictors (large-scale variables) and predictands (small-scale surface variables) do not vary under climate change conditions. The process required to adapt model outputs to end-user demands is complex.

Thus ENSEMBLES has developed a statistical downscaling portal that allows end-users to produce downscaled data without worrying about technical details. However, the portal should not be used as a black-box since this could lead to unreliable outputs or inappropriate use of downscaled data. Hence, the portal offers a support system and user guidelines.

The recommendations on the identification of robust statistical downscaling methods developed by the STARDEX project may also be of interest to users of statistical downscaling tools and outputs.

ENSEMBLES downscaling portal

The ENSEMBLES downscaling portal for statistical downscaling provides user-friendly web access to statistical downscaling techniques and simulations produced in ENSEMBLES. It allows users to choose a method of statistical downscaling and produce high-resolution predictions either using as predictands (target data), observational datasets already mounted on the server or data uploaded by the user. In this portal, GCM forecasts (seasonal-to-decadal and climate change) are downscaled to local stations or uniform observation grids using any of the available downscaling algorithms. This process is performed from a web browser following three steps: predictor selection, predictand selection and downscaling method. The portal also includes a data access tool for reanalysis, GCM and observed data sets.

The portal allows downscaling over Europe of seasonal-to-decadal hindcasts (from DEMETER and ENSEMBLES) and anthropogenic climate change simulations (from ENSEMBLES). It will be extended to West Africa in collaboration with the AMMA project.

ENSEMBLES deliverables	Details	Access
D2B.4	A first prototype of web service for downscaling at seasonal-to-decadal timescales	Public
D2B.13	ERA-40 based predictor data set for statistical downscaling	Public
D2B.17	GCM-based predictor data set for statistical downscaling	Public
D2B.19	Extension of the ENSEMBLES web-based service for downscaling	Public
D2B.23	Journal paper on a test case application of the downscaling portal for seasonal forecasts in agriculture or energy sectors	Public

Statistical Downscaling activities in ENSEMBLES

As well as developing a downscaling tool, ENSEMBLES partners have addressed the challenges of applying statistical downcaling in a probabilistic framework

ENSEMBLES deliverables	Details	Access
D2B.5	Methodology for Markov chain modelling of sequences of atmospheric circulation patterns for implementation with a conditional model of extreme hydrometeorological events	Public
D2B.12	Recommendations for the application of statistical downscaling methods to seasonal-to-decadal hindcasts in ENSEMBLES	Public
D2B.14	Recommendations for the modification of statistical downscaling methods for the construction of probabilistic projections	Public
D2B.16	Improved conditional weather generator for extreme precipitation events	Public
D2B.18	Technical protocol for the construction of ENSEMBLES statistical downscaling and scenario generator tools	Public

Non-ENSEMBLES statistical downscaling tools

Two non-ENSEMBLES statistical downscaling tools are the statistical downscaling model (SDSM) developed in the UK (Wilby, R.L., Dawson, C.W. and Barrow, E.M., 2002: 'SDSM - a decision support tool for the assessment of regional climate change impacts'. Environmental Modelling Software, 17, 145-157) and the automated statistical downscaling (ASD) tool developed in Canada. A package of R functions for statistical downscaling - clim.pact - has also been developed.

SDSM	SDSM calculates statistical relationships, based on multiple linear regression techniques, between large-scale (the predictors) and local (the predictand) climate. The downscaling models are calibrated using NCEP Reanalysis as large-scale predictors, and predictors are also provided for a number of GCM climate change simulations, including HadCM3. Version 4.1 was released in April 2007.
CCSN	SDSM provided the starting point for development of the ASD tool by CCSN in Canada.
clim.pact	A software package for retrieving data, making climate analysis and statistical downscaling of monthly mean and daily mean global climate scenarios written by Rasmus Benestad, Deliang Chen and Inger Hanssen-Bauer. This package uses the free and open source data analysis environment R and is supported by a detailed compendium.

return to top