Opened 2 years ago

Closed 12 months ago

#834 closed defect (fixed)

contfrac=0 in regional run

Reported by: jgipsl Owned by: somebody
Priority: major Milestone:
Component: Anthropogenic processes Version:
Keywords: Cc:

Description

In some regional runs with OOL_SEC_STO_FG2 (CRUJRA/v2.1/twodeg and old driver), contfrac=0 for some land pixels. How can this happen ? A grid cell with contfrac=0 should not be considered as a land pixel. Note that the model doesn't crash.

For example 3 grid celles have contfrac=0 with following domain:

LIMIT_WEST = -20.
LIMIT_NORTH = 20.
LIMIT_SOUTH = -40.
LIMIT_EAST = 50.

Also this smaller domain have some grid cells with contfrac=0:

LIMIT_WEST=22
LIMIT_NORTH=70
LIMIT_SOUTH=-30
LIMIT_EAST=32

Due to this problem, a hack has been done for the calculation of aveget where a division of contfrac was done [7529]. See also ticket #831

Change History (3)

comment:1 Changed 2 years ago by jgipsl

Test to be done if contfrac=0 also with the new driver.

comment:2 Changed 12 months ago by xnwang

If contfrac=0 exists in the forcing file, the new driver excludes such data point from the simulation, by not counting it as a valid land pixel. So, there is no such concern when using the new driver. This ticket can be closed.

comment:3 Changed 12 months ago by xnwang

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.