Evaluation of EURO 4 M Reanalysis data using

Evaluation of EURO 4 M Reanalysis data using Satellite Data Jörg Trentmann, Uwe Pfeifroth, Jennifer Lenhardt, Richard Müller Deutscher Wetterdienst (DWD) April 2013 1

General Concept • Compare monthly means from reanalysis with gridded data • So far, we used SMHI Reanalysis (1990 to 1995) (+ ERA-I for comparison) • Calculate spatial differences, distributions, correlations • Compare Anomalies, trends…. • Cloud Fraction • Surface Solar Radiation • Precipitation • Surface Albedo • Integrated Water Vapor April 2013 2

General Concept • 6 -hourly SMHI reanalysis data temporally averaged to monthly means on the 0. 2 deg rotated grid (SMHI, 1990 to 1995); for radiation + precipitation the 24 h minus 12 h forecasted accumulation was used • Regridding of all data sets (sat, reanalysis) to a common regular lon -lat grid (conservative remapping); spatial resolution 0. 2 or 0. 5 deg (determined by the lowest-resolution data set) • Generation of one file with all available monthly means and the corresponding differences • Preparation of standardized figures for comparison April 2013 3

Cloud Fraction; July 1994 SMHI – EURO 4 M Sat SMHI – ERA-I- ERA-I April 2013 EURO 4 M 4

Cloud Fraction; July (mean, 1990 – 1995) SMHI – Sat. Daten SMHI Reanalysis April 2013 5

Cloud Fraction • SMHI Reanalysis underestimates cloud Cloud Fraction; January fraction in the Mediterranean • Overestimation in January along the SMHI Reanalysis SMHI – Sat. Data Norwegian Coast April 2013 6

Cloud Fraction; Comparison with SYNOP EURO 4 M Sat. Data SMHI Reanalysis • SMHI Reanalysis fits perfectly with SYNOP Cloud Fraction, satellite data overestimates SYNOP • Cloud Fraction is a tricky parameter for comparison, because of different definitions etc April 2013 7

Solar Radiation Surface Solar Radiation; July SMHI Reanalysis SMHI – Sat. Daten April 2013 8

Surface Solar Radiation; July SMHI Reanalyse ERA-Interim ERA-I – Sat. Daten April 2013 9

Solar Radiation; Mean SMHI Reanalysis ERA-Interim • SMHI Reanalysis overestimates surface solar radiation • ERA-I compares better with Satellite Data • Interannual Variability captured by Reanalyses ERA – SMHI – Sat- Daten April 2013 10

Precipitation, Mean SMHI Reanalysis SMHI – EURO 4 M Data. Set ERA – EURO 4 M Data. Set April 2013 11

Precipitation, July SMHI Reanalysis SMHI – EURO 4 M Data. Set • Reanalyses overestimate precipitation • Interannual variability captured by Reanalyses April 2013 ERA – EURO 4 M Data. Set 12

Conclusions • SMHI reanalysis compares well with SYNOP cloud fraction; underestimates satellite-derived cloud fraction in the Mediterranean • Satellite-derived Cloud Fraction not well suited for evaluation; different definitions, viewing geometries etc. • Surface Solar Radiation overestimated in SMHI Reanalysis (ERA is doing better); clouds too thin, wrong timing? ? • Year-to-year variability captured • Too much precipitation in SMHI (and ERA-I) reanalysis, especially in mountainous regions, e. g. Alps April 2013 13

Next Steps • Assess the surface albedo and integrated water vapor • Develop additional quality measures / metrics, e. g. , for the ability of the reanalysis to quantify anomalies / trends etc. • How to perform the evaluation in the “space of the user”? What means “good enough”? • How to communicate the results to possible users of reanalysis data? What are the consequences of these results? April 2013 14

April 2013