At the WMO Commission for Basic Systems (CBS) ET-OWFPS Meeting on 22-24 October 2014 in Geneva, a draft document on standardized surface verification of deterministic NWP products was created. The document specifies the content of the future exchange of surface scores between global centres and is available here: Standardized Surface Verification for Deterministic NWP
The format for the exchange of surface scores has not been defined yet. ECMWF will be collecting and archiving the scores and proposes a simple ascii format. Within this format, two options are currently under consideration.
(a) Vertical-bar (VBar) delimited format: slightly simpler to implement but less flexible with regard to possible future extensions
(b) Record format: slightly more complicated but more flexible, with the additional advantage that it is the same as the one proposed for upper-air scores
The parameter keys and lists of parameters and scores are the same in both cases.
Data file consists of score values and corresponding metadata in an ASCII format.
The first line contains a tag setting the version of the file format:
#version=1.0
There is currently only one version of the format: 1.0.
Ensuing lines contain data records. Each record has the following format:
centre | model_id | yyyymm | time | forecast_step | station_id | latitude | longitude | station_elevation | model_orography_elevation | parameter | score | event | sample_size | score_mean_value |
Data file consists of score values and corresponding metadata in an ASCII format.
Every score value is described by the full set of key attributes, like its parameter, station id, month, step etc. Parameters describing one score value at one station are organised into a record. Each record corresponds to one score value. A record is a collection of pairs key=value separated by commas. A record spans one line. Value of the key which is not given in the current record is inherited from the previous record, except for the value parameter v which has to be present in each record.
Each record has the following format:
| centre=centre, model=model_id, d=yyyymm, t=time, st=station_id, lat=latitude, lon=longitude, se=station_elevation, me=model_orography_elevation, par=parameter, sc=score, ev=event, n=sample_size, v=mean_value |
If the value is not available the record either should not be given at all or the value should be set to NIL (v=nil); every record must contain the key v (value of key v is not inheritable from previous record).
model_id (a string, not containing a comma or vertical bar) is free model identifier assigned by the originating centre (to distinguish between potentially different models provided by the centre);
| parameter | description | unit |
|---|---|---|
| t2m | air temperature at 2 meters above the model orography, corrected to the actual station elevation using the constant lapse rate 6.5K/1000m | K |
| td2m | dewpoint at 2 meters above the model orography | K |
| rh2m | relative humidity at 2 meters above the model orography | % |
| tp06 | total precipitation accumulated over previous 6 hours | mm |
| tp24 | total precipitation accumulated over previous 24 hours | mm |
| ff10m | wind speed at 10 meters above the model orography | m/s |
| dd10m | wind direction at 10 meters above the model orography | deg |
| tcc | total cloud cover | 0-1 (convert to okta for contingency table) |
| score | description | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| me | mean error (bias) | |||||||||
| mae | mean absolute error | |||||||||
| rmse | root mean square error | |||||||||
| ct | contingency table, the 4 values are misses, hits, correct non-events, and false alarms (in this order)
|
| event | description |
|---|---|
| val>thr | forecast/observed value greater than a threshold value of the forecast parameter, e.g. for wind speed >15 m/s: val>15 |
| val<=thr | forecast/observed value smaller than or equal to a threshold value of the forecast parameter, e.g. for cloudiness of 0-2 okta: val<=2 |