Download presentation
Presentation is loading. Please wait.
Published byTracy Blake Modified over 9 years ago
1
Federal Department of Home Affairs FDHA Federal Office of Meteorology and Climatology MeteoSwiss Postprocessing of temperature and wind for COSMO-7 and COSMO-2 Vanessa Stauch Offenbach, September 2009 COSMO General Meeting
2
2 Statistical PP | COSMO-GM 2009 Vanessa Stauch calibration with Kalman Filter >> recursive estimation of forecast error (prediction – correction) >> requires online observations >> can be used quasi-instantaneously (no large historical database) >> cannot predict fast changes (assumption of persistent error) >> suitable for a subset of parameters (normally distributed errors)
3
3 Statistical PP | COSMO-GM 2009 Vanessa Stauch calibration with Kalman Filter error model: states evolution: prediction: with error : ^ ^
4
4 Statistical PP | COSMO-GM 2009 Vanessa Stauch COSMO models COSMO-7 COSMO-LEPS COSMO-2 COSMO-LEPS 10km, +132 hours COSMO-7 6.6km, +72 hours COSMO-2 2.2km, +24 hours
5
5 Statistical PP | COSMO-GM 2009 Vanessa Stauch Kalman Filter @ MeteoSwiss operational: T2m, TD2m for COSMO-LEPS mean COSMO-7 COSMO-2 IFS in preparation: FF10m, TW2m, RH2m for COSMO-LEPS mean COSMO-7 COSMO-2 IFS
6
6 Statistical PP | COSMO-GM 2009 Vanessa Stauch Swiss met. measurement network 62 stations
7
7 Statistical PP | COSMO-GM 2009 Vanessa Stauch T2m predictions COSMO-7 COSMO-2 KF C7 COSMO-7 KF C2 COSMO-7 performance?
8
8 Statistical PP | COSMO-GM 2009 Vanessa Stauch benefit COSMO-2 vs COSMO-7? 04.04.08 – 31.10.08 RMSE All ANETZ stations 11.0 %4.0 % Low ANETZ stations 8.8 %5.2 % High ANETZ stations 13.0 %3.2 % 04.04.08 – 31.10.08 STD All ANETZ stations 12.3 %4.0 % Low ANETZ stations 12.7 %5.2 % High ANETZ stations 12.6 %3.2 % C2 vs C7C2-KF vs C7-KF =
9
9 Statistical PP | COSMO-GM 2009 Vanessa Stauch benefit COSMO-2 vs COSMO-7-KF?? 04.04.08 – 31.10.08 RMSE C2-DMO vs C7-KF STD C2-DMO vs C7-KF All ANETZ stations -25 %-20 % Low ANETZ stations -22 %-12 % High ANETZ stations -27 %-20 %
10
10 Statistical PP | COSMO-GM 2009 Vanessa Stauch Chasseral (CHA) Evionnaz (EVI) Gütsch (GUE) Piz Martegnas (PMA) Schaffhausen (SHA) Oron (ORO) Üetliberg (UEB) stations for wind speed calibration SMN station WKA
11
11 Statistical PP | COSMO-GM 2009 Vanessa Stauch StationCHAEVIGUEOROPMASHAUEBWiColWiFelWiGueWiCro Höhe (obs)1599480228783026704371043450102023311230 cosmo7WiCroWiColWiGueOROPMASHAUEBWiColWiFelWiGueWiCro Höhe (mod)10881024232281123344325411024101323221088 deltaheight_7 (mod-obs) -51154435-19-336-4.8-502574-6.8-8.8-142 cosmo2CHAEVIWiGueOROPMASHAUEBWiColWiFelWiGueWiCro Höhe (mod)12937462298805252047760479694122981114 deltaheight_2 (mod-obs) -30626611-25-15040-439346-79-33-116 CHA/WiCro EVI/WiCol WiGue PMA SHA UEB ORO WiFel height differences
12
12 Statistical PP | COSMO-GM 2009 Vanessa Stauch represenativeness of met. station model prediction representative for (mean) grid box local point observation (specific conditions) wind turbine Gütsch
13
13 Statistical PP | COSMO-GM 2009 Vanessa Stauch COSMO-7 vs COSMO-2 COSMO-7COSMO-2 (03) Chasseral (CHA) 56>44 Evionnaz (EVI) 90<99 Gütsch (GUE) 55>45 Oron (ORO) 53<59 Piz Martegnas (PMA) 51>45 Schaffhausen (SHA) 54>52 Uetliberg (UEB) 76>69 rRMSE (%) für 1-24h, period 01.09.08 – 31.03.09 CHA EVI GUE PMA SHA ORO UEB
14
14 Statistical PP | COSMO-GM 2009 Vanessa Stauch effect on MOS-postprocessing COSMO-7 MOS COSMO-2 (03) MOS Chasseral (CHA) 34>32 Evionnaz (EVI) 78>77 Gütsch (GUE) 45>39 Oron (ORO) 59>47 Piz Martegnas (PMA) 69>42 Schaffhausen (SHA) 85>77 Uetliberg (UEB) 59>54 rRMSE (%) für 1-24h, period 01.09.08 – 31.03.09 CHA EVI GUE PMA SHA ORO UEB
15
15 Statistical PP | COSMO-GM 2009 Vanessa Stauch effect on KF-postprocessing COSMO-7 KF COSMO-2 (03) KF Chasseral (CHA) 33= Evionnaz (EVI) 77<83 Gütsch (GUE) 47>43 Oron (ORO) 52>50 Piz Martegnas (PMA) 48>43 Schaffhausen (SHA) 44= Uetliberg (UEB) 57>56 rRMSE (%) für 1-24h, Zeitraum 01.09.08 – 31.03.09 CHA EVI GUE PMA SHA ORO UEB
16
16 Statistical PP | COSMO-GM 2009 Vanessa Stauch summary >> statistical postprocessing profits from a better NWP input model >> „dynamical downscaling“ does not replace statistical adaptation to local observations (in particular if results being verified against those)
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.