Opened 8 years ago

Last modified 8 years ago

#83 accepted task

Prevent cold pool formation

Reported by: raw88 Owned by: raw88
Priority: normal Milestone:
Component: Atmosphere Version:
Keywords: Cc:

Description

Edit to routine lsp_evap.F90 in atmosphere/large_scale_precipitation to remove evaporation of rain. Purpose is to prevent cold pool formation under convective storms in a case study simulation. Added line dpr(i) = dpr(i)*0.001 after section which calculates transfers.

Change History (4)

comment:1 Changed 8 years ago by raw88

  • Owner set to raw88
  • Status changed from new to accepted

comment:2 Changed 8 years ago by raw88

  • Summary changed from Switch off rain evaporation to Prevent cold pool formation

Switching off rain evaporation did not completely remove cold pools. Will now try switching off melting of ice & aggregates and evaporation of melting ice & aggregates. This will done by making the same edit described above in the routines lsp_evap_snow.F90 and lsp_melting.F90.

comment:3 Changed 8 years ago by raw88

Switching off melting of aggregates & crystals and evaporation of melting aggregates & crystals still did not eliminate cold pools. Part of the reason appears to be my failing to edit the routine lsp_tidy.F90 which evaporates left over ice at temperatures above 0 deg C. Thus melting of ice was still occuring regardless of my edits. I will now try a new approach - rather than switching off the evaporation and melting terms completely I will simply prevent these processes from altering the temperature field by commenting out the lines

T(i) = T(i) - lsrcp * dpr(i)

in the routines lsp_evap.F90, lsp_melting.F90, lsp_evap_snow.F90, and lsp_tidy.F90.

comment:4 Changed 8 years ago by raw88

This second method worked - the temperature changes associated with phase changes of water below the freezing level were removed, thereby preventing cold-pool formation. However, the effect on the simulation was minor, suggesting that cold pools were not very strong anyway in this case.

To test the effect of stronger cold pools a new run will be carried out, where the temperature changes in large-scale precipitation routines listed above are switched back on and doubled. That is,

T(i) = T(i) - lsrcp * dpr(i)

becomes

T(i) = T(i) - lsrcp * 2 * dpr(i)

Note: See TracTickets for help on using tickets.