Methodology and user guide

Overview

This app uses data made available by the Copernicus Climate Change Service (C3S) Climate Data Store. C3S is part of the Copernicus Earth Observation Programme and implemented by the European Centre for Medium-Range Weather Forecasts (ECMWF) on behalf of the European Commission.

The app allows users to view and analyze the European Centre for Medium-Range Weather Forecasts (ECMWF) Total Precipitation (rainfall) forecasts. It also allows users to explore how skilful these forecasts are in relation to predicting potential for upcoming droughts based on historical performance.

Its target users are UK water companies. Water companies are primarily interested in exploring the ECMWF data and its potential for forecasting upcoming drought. Their primary means of categorising drought is in relation to Long Term Average (LTA) rainfall, derived from the CEH GEAR data (see below).

Units of measurement

Units of measurement for rainfall across all datasets have been standardised to mm per day.

All datasets have been aggregated to ECMWF's 1 degree by 1 degree spatial grid.

All datasets have been aggregated to monthly temporal bins.

Historical observed rainfall data

Data from CEH-GEAR, derived from Met Office national database of observed precipitation, is used as the historical observed dataset. This is used for the verification of the skills of ECMWF hindcasts and to derive the Long Term Average rainfall. Gridded estimates of daily and monthly areal rainfall for the United Kingdom (1890-2017) [CEH-GEAR]

Long Term Average rainfall data

Derived from Met Office national database of observed precipitation, CEH-GEAR. The observed average rainfall between the years of 1961 and 1991 is taken as the Long Term Average and is a benchmark used by UK water companies and the Enviornment Agency. Gridded estimates of daily and monthly areal rainfall for the United Kingdom (1890-2017) [CEH-GEAR]

Recently observed rainfall data

Recently observed rainfall data is from the ECMWF ERA5 dataset linked below. This is used to display the previous months observed rainfall data on the main graph. ERA5 monthly averaged data on single levels from 1979 to present

Upcoming forecast data

Upcoming forecast data is from the ECMWF SEAS dataset linked below. This uses an ensemble of 51 forecasts for each grid. The mean of these forecasts is taken as the most likely forecast. Seasonal forecast monthly statistics on single levels

Historical skill

Historical skill of hindcasts is carried out per grid, month and leadtime and spans 23 years from 1994 to 2016. Historical skill calculations are based upon how well the hindcast ensembles in the historic period performed in correctly predicting lower rainfall thresholds (80% and 60% of Long Term Average). Seasonal forecast daily data on single levels

Bias correction

Anomaly bias correction of all ensemble members in the forecasts and hindcasts was carried out in order to correct the model bias in line with observations.

Disclaimer

Forecasting estimations of rainfall at this range and scale has potentially large margins for error. The ECMWF and H.R. Wallingford Ltd take no responsibility for any losses or damages incurred by anyone using this software or any of the data held within it.