The D-Phase MicroPEPS is a LAF-Ensemble (lagged average forecast) that is based upon 5 different high resolution models: COSMOCH2 LMK AROME CMCGEMH ISACMOL2 The individual members can be found in the respective experiments 'dphase_*' in this data base. As time lagged forecasts the initialisation times t-3h and t-6h relative to the reference time t are incorporated. In a maximum the MicroPEPS might have 9 members (COSMOCH2: 3, LMK: 3, AROME: 1, CMCGEMH: 1, ISACMOL: 1). While COSMOCH2 and LMK generate new forecasts every 3 hours the other models run less frequently. In the 6 hour MicroPEPS time window AROME, CMCGEMH and ISACMOL provide one run each. During operation the ensemble size might change due to the availability of the forecasts. The MicroPEPS uses equal weights for averaging its members. The MicroPEPS generates probability forecasts by interpreting the overlapping areas of the single forecasts as members of a local ensemble. Due to the different domains of the deterministic models the size of the ensemble depends on location. Hence the quality of the forecasted probability distributions varies over the domain. There will be four runs a day at 0, 6, 12 and 18 UTC. Grid description: CDOM: xfirst: 6.0 yfirst: 47.0 xsize: 168.0 ysize: 151.0 xinc: 0.03 yinc: 0.02 xnpole: 0.0 ynpole: 0.0 DDOM: xfirst: 2.0 yfirst: 43.0 xsize: 535.0 ysize: 351.0 xinc: 0.03 yinc: 0.02 xnpole: 0.0 ynpole: 0.0

The SRNWP-PEPS consists of 21 different operational limited area models: Weather Service / Limited Area Model / Resolution [km] / Coupling Model / Forcast Period / Time Interval [h] / Main Run [UTC] Denmark HIRLAM 16 ECMWF +60h 1 0, 6, 12, 18 Finland HIRLAM 22 ECMWF +54h 1 0, 6, 12, 18 Ireland HIRLAM 16 ECMWF +48h 3 0, 6, 12, 18 Netherlands HIRLAM 22 ECMWF +48h 1 0, 6, 12, 18 Spain HIRLAM 18 ECMWF +48h 1 0, 6, 12, 18 Norway HIRLAM 22 ECMWF +30h 1 0, 12 Sweden I HIRLAM 11 ECMWF +48h 3 0, 6, 12, 18 Sweden II HIRLAM 22 ECMWF +48h 3 0, 6, 12, 18 Belgium ALADIN 15 ARPEGE +60h 1 0, 12 Austria ALADIN_A 9.6 ARPEGE +48h 1 0, 12 France ALADIN_F 11 ARPEGE +48h 3 0, 12 Croatia ALADIN_L 8.9 ARPEGE +48h 3 0, 12 Czech Rep. ALADIN_L 11 ARPEGE +48h 1 0, 12 Hungary ALADIN_L 11 ARPEGE +48h 1 0, 12 Slovakia ALADIN_L 11 ARPEGE +48h 3 0, 12 Slovenia ALADIN_L 9.5 ARPEGE +48h 3 0, 12 United Kingdom UKMO-LAM 12 UM global +48h 3 0, 6, 12, 18 Germany LME 7 GME +78h 1 0, 12 // Germany LME 7 GME +48h 1 6, 18 Switzerland aLMo 7 ECMWF +72h 1 0, 12 Italy EuroLM 7 EuroHRM +60h 3 0, 12 Poland LM 14 GME +72h 3 0, 12 The relation between these models and the numbers of the SRNWP PEPS ensemble is anonymous. The SRNWP-PEPS generates probability forecasts by interpreting the overlapping areas of the single forecasts as members of a local ensemble. Due to the different domains of the deterministic models the size of the ensemble depends on location. Hence the quality of the forecasted probability distributions varies over the PEPS domain. THE PLEV data set could not be provided for PEPS. Grid description: xsize,ysize,xinc, yinc differ for ensemble members (see datasets) CDOM: xfirst: 6.0 yfirst: 47.0 xnpole: 0.0 ynpole: 0.0 DDOM: xfirst: 2.0 yfirst: 43.0 xnpole: 0.0 ynpole: 0.0

This experiment contains forecasts from the LME (COSMO-EU) model of DWD (7km horizontal resolution, 40 model levels). Model runs are started every 6h at 00, 06, 12 and 18 UTC with a forecast range of +72h. LME (COSMO-EU) is an operational forecast model of DWD. The output is mostly according to the tigge+ list. Descriptions of the differences to TIGGE+ can be found in the summary of the data sets. For a detailed description of the LME (COSMO-EU) model, please contact the originator of the data. Grid description: CDOM: xfirst: -2.73 yfirst: -2.927 xsize: 177.0 ysize: 112.0 xinc: 0.063 yinc: 0.063 xnpole: -170.0 ynpole: 40.0 DDOM: xfirst: -5.882 yfirst: -6.685 xsize: 177.0 ysize: 112.0 xinc: 0.063 yinc: 0.063 xnpole: -170.0 ynpole: 40.0

The new version of the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data set - HOAPS II - contains improved global fields of precipitation and evaporation over the oceans and all basic state variables needed for the derivation of the turbulent fluxes. Except for the NOAA Pathfinder SST data set, all variables are derived from SSM/I satellite data over the ice free oceans between 1987 and 2002. The earlier HOAPS version was improved and includes now the utilisation of multi-satellite averages with proper inter-satellite calibration, improved algorithms and a new ice detection procedure, resulting in more homogeneous and reliable spatial and temporal fields as before. Here, 5-days mean( Pentad )datasets are available but monthly mean and climatological means are also public and available via the CERA database system. The spatial resolution of 0.5 degree, makes them ideally suited for studies of climate variability over the global oceans. Further information under : http://www.hoaps.zmaw.de .

The new version of the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data set - HOAPS II - contains improved global fields of precipitation and evaporation over the oceans and all basic state variables needed for the derivation of the turbulent fluxes. Except for the NOAA Pathfinder SST data set, all variables are derived from SSM/I satellite data over the ice free oceans between 1987 and 2002. The earlier HOAPS version was improved and includes now the utilisation of multi-satellite averages with proper inter-satellite calibration, improved algorithms and a new ice detection procedure, resulting in more homogeneous and reliable spatial and temporal fields as before. The spatial resolution of 0.5 degree, makes them ideally suited for studies of climate variability over the global oceans. Pentade and climatological means are also public and available via the CERA database system. Further information under : http://www.hoaps.zmaw.de .

The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) set is a completely satellite based climatology of precipitation, turbulent heat fluxes and freshwater budget (evaporation minus precipitation) as well as related athmospheric state variables over the global ice free oceans. All variables are derived from SSM/I passive microwave radiometers, except for the SST, which is taken from AVHRR measurements. The dataset includes multi-satellite averages, inter-sensor calibration, and an efficient sea ice detection procedure. Changes in this version are a longer time series, now containing data from 1987 to 2005, a new neural network based precipitation algorithm, and inclusion of the RSMAS/NODC Pathfinder Version 5 SST fields. Additionally a new 85 GHz synthesis procedure has been implemented, making a continuous time series for all parameters for the whole time series possible. This dataset contains 1 degree twice daily globally gridded multi-satellite composite products, providing high temporal resolution. Each grid-cell contains data from only one satellite pass, there is no average from two or more satellites. Early passes are overwritten by later passes. This method provides more spatial homogeneity than averaging all available data. The fields are stored for 0-12 and 12-24 UTC. Timesteps in the data files are at 0 UTC (0-12 UTC overpasses) and 12 UTC (12-24 UTC overpasses). Each grid-cell contains the average of data from the satellite that passed this gridbox closest to 12 and 24 UTC, respectively. Other gridded data sets available are pentad (5-day) and monthly means on a global 0.5 deg. x 0.5 deg. grid. For more information see http://www.hoaps.org/.

The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) set is a completely satellite based climatology of precipitation, turbulent heat fluxes and freshwater budget (evaporation minus precipitation) as well as related athmospheric state variables over the global ice free oceans. All variables are derived from SSM/I passive microwave radiometers, except for the SST, which is taken from AVHRR measurements. The dataset includes multi-satellite averages, inter-sensor calibration, and an efficient sea ice detection procedure. Changes in this version are a longer time series, now containing data from 1987 to 2005, a new neural network based precipitation algorithm, and inclusion of the RSMAS/NODC Pathfinder Version 5 SST fields. Additionally a new 85 GHz synthesis procedure has been implemented, making a continuous time series for all parameters for the whole time series possible. All pentad (5-day) mean products are distributed in separate monthly files. The first day of a 5-day average period determines the monthly file wherein it is saved. Each year is subdivided in 73 pentads starting at the same day of the year. During leap years, the twelfth pentad (starting at February 25) is the average of six days. Apart from pentad (5-day) and monthly means on a global 0.5 deg. x 0.5 deg. grid, twice daily multi-satellite composite data on a global 1 x 1 deg. grid are available. For more information see http://www.hoaps.org/.

The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) set is a completely satellite based climatology of precipitation, evaporation and freshwater budget (evaporation minus precipitation) as well as related turbulent heat fluxes and atmospheric state variables over the global ice free oceans. All variables are derived from SSM/I passive microwave radiometers, except for the SST, which is taken from AVHRR measurements. The data set includes multi-satellite averages, inter-sensor calibration, and an efficient sea ice detection procedure. Changes in this version are a prolonged time series, now containing data from 1987 to 2005, a new neural network based precipitation algorithm, and inclusion of the RSMAS/NODC Pathfinder Version 5 SST fields. Additionally a new 85 GHz synthesis procedure has been implemented for the time period to compensate for the missing channel information on DMSP F08, see accuracy report. Apart from monthly and pentad (5-day) means on a global 0.5 deg. x 0.5 deg. grid, twice daily multi-satellite composite data on a global 1 x 1 grid are available.

This experiment contains forecasts from the LMK (COSMO-DE) high resolution model of DWD (2.8km horizontal resoultion and 50 model levels). Model runs are started every 3h at 00, 03, 06, 09, 12, 15, 18 and 21 UTC with a forecast range of +18h. LMK (COSMO-DE) is an operational forecast model of DWD. Therefore, we adapted the output of the model as close as possible to the tigge+ list, but there are some differences; see dataset summaries. For a detailed description of the LMK (COSMO-DE) model, please contact the originator of the data. All datasets for COPS in the database have an output frequency of 15 minutes. If the variables are not provided by LMK (COSMO-DE) with an output frequency of 15 minutes then the hourly output has been linearily interpolated in time. LMK (COSMO-DE) provides only a subset of the TIGGE+ variables with an output frequency of 15 minutes. These are: Total precipitation (all types) (kg/m**2) acc_st 011 002 TPT2 Precipitation: grid-scale only, rain (kg/m**2) acc_st 102 201 SURF Precipitation: grid-scale only, snow (kg/m**2) acc_st 079 002 SURF Precipitation: grid-scale only, graupel (kg/m**2) acc_st 132 201 SURF Precipitation rate: grid-scale only, rain (kg/s/m**2) inst 100 201 SURF Precipitation rate: grid-scale only, snow (kg/s/m**2) inst 100 201 SURF Precipitation rate: grid-scale only, graupel (kg/s/m**2) inst 100 201 SURF Total column water vapour (or precipitable water) (kg/m**2) inst 054 002 SURF Total column cloud water (or cloud water) (kg/m**2) inst 076 002 SURF Total column cloud ice (or cloud ice) (kg/m**2) inst 058 002 SURF W-velocity (m/s) inst 040 002 MUVW Grid descitption: CDOM: xfirst: -2.73 yfirst: -2.927 xsize: 135.0 ysize: 118.0 xinc: 0.025 yinc: 0.025 xnpole: -170.0 ynpole: 40.0 DDOM: xfirst: -5.882 yfirst: -6.685 xsize: 441.0 ysize: 279.0 xinc: 0.025 yinc: 0.025 xnpole: -170.0 ynpole: 40.0