(more details francis.pouponneau++at++meteo.fr )
Next tables are cloud cover and precipitation contingency tables for the 36H forecast range of ALADIN-France model (0 hours initial data set). Classes are those used in the verification of NWP models in EUROPE (EWGLAM). Contingency tables are here computed over France.
Three skill scores have been calculated over the whole table :
Contingency tables for 6 hours precipitation periods are computed with classes :
In contingency tables, blue color means an underestimation, red one represents an overestimation.
NEBULOSITY
|
9810-9812 |
CLOUD COVER |
||||
|
MODEL : |
ALADIN |
00 H start | |||
|
AREA : |
FRANP01-Franp01 |
Forecast
Range
|
36 H | ||
|
92 days |
|||||
Forecast |
|||||
0 - 2 |
3 - 6 |
7 - 8 |
SUM | ||
0 - 2 |
11.3%
|
6.8%
|
0.2%
|
18.3%
| |
Obs. |
3 - 6 |
10.2%
|
17.9%
|
4.0%
|
32.2%
|
7 - 8 |
6.5%
|
27.5%
|
15.5%
|
49.5%
| |
SUM |
27.9%
|
52.3%
|
19.8%
|
15982
| |
|
Correct
:
|
44.7% |
Rousseau
:
|
0.15 |
Heidke
:
|
0.19 |
The percentage of correct is 44.7%. The high nebulosity class (7-8) is more observed than forecasted : 49.5% against 19.8% . Heidke and Rousseau indexes are low (0.15 et 0.19).
PRECIPITATIONS
|
9810-9812 |
PRECIPITATIONS |
|||||
|
MODEL : |
ALADIN |
00 H start |
||||
|
AREA : |
FRANP01-Franp01 |
Forecast
Range
|
36 H |
|||
|
92 days |
||||||
Forecast |
||||||
NO rain |
LIGHT |
MODERATE |
HEAVY |
SUM | ||
NO rain |
61.2%
|
8.8%
|
1.9%
|
0.2%
|
72.1%
| |
LIGHT |
8.7%
|
6.6%
|
2.7%
|
0.2%
|
18.3%
| |
Obs. |
MODERATE |
1.9%
|
3.5%
|
2.9%
|
0.5%
|
8.7%
|
HEAVY |
0.1%
|
0.2%
|
0.4%
|
0.2%
|
0.9%
| |
SUM |
71.9%
|
19.2%
|
7.8%
|
1.1%
|
17994
| |
|
Correct
:
|
70.9% |
Rousseau
:
|
0.34 |
Heidke
:
|
0.34 | |
The percentage of correct is more than 70.9%. We can see that there is as often "light rain" forecasted whereas "no rain" was observed as the reverse. We don't find again the tendency described in previous Newsletter : a shift between "no rain" and "light rain". If we consider only two classes "rain" and "no rain", again there are as many cases where "rain" was predicted whereas "no rain" was observed as the reverse.
|
Home |