Special efforts are put in the validation of MetGIS snowfall forecasts. However, this is not an easy task since measurements of fresh snow may be quite faulty, especially when snow falls in combination with the occurrence of strong winds.
Automated weather stations usually exhibit large errors and inconsistencies in their fresh snow measurements, and sometimes it may be difficult to judge if the deviation between measurements and predictions is due to observation errors or caused by forecast deficiencies. Therefore numbers are presented (see the adjacent verification table) that stem from a study that includes a Chilean mining station in the Andes mountains where human observers measure fresh snow in hourly intervals. The displayed table offers information about snowfall observed and forecast by MetGIS at the Los Pelambres site (3185 m) for the most important snowstorms of the 2009 winter season. The results demonstrate that MetGIS is an excellent tool for decent snowfall forecasts.
|Snowfall at Los Pelambres||Depth of observed |
fresh snow [cm]
|Depth of fresh snow |
forecast by MetGIS [cm]
|From (start time)||Until (end time)|
|2009-05-29 09:00||2009-05-30 00:00||9||6|
|2009-06-28 04:30||2009-06-30 03:00||103||87|
|2009-07-14 11:00||2009-07-15 07:00||12||21|
|2009-07-21 07:00||2009-07-22 13:00||55||53|
|2009-07-31 23:00||2009-08-01 04:00||4||1|
|2009-08-15 03:00||2009-08-16 04:00||51||89|
|2009-08-17 19:00||2009-08-19 11:00||44||92|
|2009-09-06 10:00||2009-09-06 23:30||4||3|
|2009-09-07 03:30||2009-09-07 12:00||6||5|
Regarding the snow line it has to be said that it is very hard to validate this parameter, because it usually does not get observed. Experiments focusing on the occurrence of snowfall at alpine stations in Europe showed that the MetGIS snow line estimates are quite accurate in general, but tend to give a higher snow line than observed in inner alpine heavy precipitation events. An improvement of forecasts for situations like this is topic of ongoing work, and will soon be implemented.