1. ABOUT THE DATASET ------------ Title: Post-processed reforecasts of the European Flood Awareness System and related evaluation data Creator(s): Gwyneth Matthews[1], Christopher Barnard[2] Contributor(s): Hannah Cloke[1,2,3,4], Sarah Dance[1], Toni Jurlina [2], Cinzia Mazzetti[2], Christel Prudhomme[2,6,7] Organisation(s): 1.University of Reading. 2.European Centre for Medium-range Weather Forecasts. 3.Uppsala University. 4.Centre of Natural Hazards and Disaster Science. 6.University of Loughborough. 7.UK Centre for Ecology and Hydrology. Rights-holder(s): European Centre for Medium-range Weather Forecasts Publication Year: 2021 Description: This dataset contains the twice-weekly post-processed reforecasts for years 2017 and 2018. These reforecasts were created by post-processing the Copernicus Emergency Management Service's European Flood Awareness System (EFAS) reforecasts. The post-processing method used is the operational method for the EFAS (Oct 2021). These post-processed reforecasts were used in the evaluation of the post-processing method. Additional data used in the evaluation is also included. The python code used in the evaluation is provided. The python scripts contain functions for the calculation of verification metrics for both the post-processed and the non-post-processed reforecasts. Cite as: Matthews, Gwyneth and Barnard, Christopher (2021): Post-processed reforecasts of the European Flood Awareness System and related evaluation data. University of Reading. Dataset. http://dx.doi.org/10.17864/1947.333. Related publication: Matthews, Barnard, Cloke, Dance, Jurlina, Mazzetti, and Prudhomme. Evaluating the impact of post-processing medium-range ensemble streamflow forecasts from the European Flood Awareness System. HESS. 2021. (In preparation). Contact: g.r.matthews@pgr.reading.ac.uk 2. TERMS OF USE ----------------- Copyright. 2021. European Centre for Medium-range Weather Forecasts. This dataset is licensed under a Creative Commons Attribution 4.0 International Licence: https://creativecommons.org/licenses/by/4.0/. 3. PROJECT AND FUNDING INFORMATION ------------ Title: Multi-model data assimilation techniques for flood forecasting (PhD project). Dates: September 2019 Funding organisation: Engineering and Physical Sciences Research Council (EPSRC) Grant no.: EPSRC DTP 2018-19 University of Reading (EP/R513301/1) Funding organisation: European Centre for Medium-range Weather Forecasts Grant no.: N/A 4. CONTENTS ------------ File listing (1) data ./data/station_data.csv: station ID, stationLon: station longitude, stationLat: station latitude, regulated: 1:station is regulated, 0: station is not regulated, upstream area: 1: less that 1000km2, 2:between 1000km2 and 5000km2, 3: above 5000km2, elevation: 1: less than 150m above sea level, 2: between 150m and 400m above sea level, 3: above 400m above sea level, time of concentration: 1: less than 24 hours, 2: between 24 hours and 48 hours, 3: move than 48 hours, calibration timeseries length: 1: less than 15 years, 2: between 15 years and 20 years, 3: between 20 years and 25 years, 4: over 25 years. MQ: mean annual flow threshold (data not provided, column is required for code execution), MHQ: mean annual maximum flow threshold (data not provided, column is required for code execution), 10%: 10% threshold (data not provided, column is required for code execution) ./data/forecast_dates.txt: dates of reforecasts that are provided, %Y%m%d%H e.g. 3rd January 2017 is 2017010300 ./data/pp_reforecasts/ - individual post-processed forecasts for each station of interest. File names: ./data/pp_reforecasts//pp_reforecast__.csv File structure: Each file contains the forecast for a single station and a single date. Columns are labelled q1, q2, ...., q98, q99 and refer to the quantiles of the forecast distribution. Rows are labelled by the date of each timestep in the forecast period in the format %Y%m%d%H. Forecast units: m^3s^{-1} [cumecs] (2) Scripts ./scripts/evaluation_functions.py: Contains functions used to calculate the evaluation metrics. ./scripts/evaluation_of_forecasts.py: Contains code to evaluate each forecast and to collate the results for analysis. Calls functions from ./scripts/evaluation_functions.py. ./scripts/run_evaluation.py: Contains code to run the evaluation of the post_processed and raw reforecasts. Calls functions from ./scripts/evaluation_functions.py. 5. METHODS -------------------------- Post-processed forecasts were created using the Copernicus Emergency Management Service's EFAS v4.0 reforecasts (Barnard, C., Krzeminski, B., Mazzetti, C., Decremer, D., Carton de Wiart, C., Harrigan, S., Blick, M., Ferrario, I., Wetterhall, F., Thiemig, V., Salamon, P., Prudhomme, C. (2020). Reforecasts of river discharge and related data by the European Flood Awareness System, version 4.0, Copernicus Climate Change Service (C3S) Climate Data Store (CDS). (Accessed: 4 March 2021), 10.24381/cds.c83f560f). The post-processing method used is described in detail in Section 3 of Matthews, Bernard, Cloke, Dance, Jurlina, Mazzetti, and Prudhomme. Evaluating the impact of post-processing medium-range ensemble streamflow forecasts from the European Flood Awareness System. HESS. 2021. (In preparation). In order to run this evaluation observations must be obtained from the local authorities. The format in which observations are provided may vary by local authority. In order to run the evaluation the observations must be stored as .npy files in the Data folder with paths ./data/observations//obs__.npy. Values for the mean annual flow thresholds (MQ), the mean annual maxium flow thresholds (MHQ), and the 10% thresholds for all stations must also be calculated and added to the respective columns of ./data/station_data.csv. Once added the evaluation can be run entirely using ./scripts/run_evaluation.py.