Looking for good test cases for RR_24 redistribution algorithm

Plan:

  1. Review 24hr precipiation charts from http://senorge.no/
  2. Classify each day if it is
    1. No rainfall
    2. Convective Rainfall
    3. Frontal/uniform precipitation
  3. If possible do this based on regions

Initial test regions:

Later will iterate to better region split based on learning … where is that AI neural net when you need it ? ;-)


Suggestion: Precipitation regions is used for rainfall and for runoff. Data for these regions can be used to make trends in accumulated precipitation, see met.no - the trend diagram here is made for geographical regions, but the same has been done for the precipitation regions.

Precipitation regions:

nedborregionkart.jpg

The attached file identifies each station with its particular precipitation region (number in the NR_NO nedbør region column) as illustrated by the following excerpt:

STNR UTM_E UTM_N GR_NO TR_NO NR_NO
900 337498 6926514 4 1 7
10300 333750 6930672 4 1 7
10380 312615 6943795 4 1 7
10400 314190 6942590 4 1 7
10430 310528 6953895 4 1 7
10600 318477 6954240 4 1 7
10650 337833 6956155 4 1 7
10700 338755 6961042 4 1 7
10740 340079 6949931 4 1 7
10750 340289 6950456 4 1 7
10800 337048 6953941 4 1 7