Opened 7 years ago

Closed 6 years ago

#137 closed defect (fixed)

possible causes of instability by 'pdqsold' in 'enerbil_begin'

Reported by: fwang Owned by: fwang
Priority: major Milestone:
Component: Anthropogenic processes Version: trunc
Keywords: instability, pdqsold Cc:

Description

  1. Location of the code:

model version: 1347

  SUBROUTINE enerbil_begin
  !! 3. Computes pdqsold
pdqsold(ji) = dev_qsol(ji) * ( pb(ji)**kappa ) / cp_air
  1. Problem description

(1) '1/cp_air' term is already included during calculating the sensitivity terms (zicp = un/cp_air in SUBROUTINE enerbil_surftemp). e.g.,:

larsub_sns = chalsu0 * vbeta1(ji) * (un - vbeta5(ji)) * zicp(ji) *
pdqsold(ji) / (zikq - peqAcoef(ji))   
lareva_sns = chalev0 * ((un - vbeta1(ji))*(un - vbeta5(ji)) * vbeta(ji) * valpha(ji) + vbeta5(ji)) * &
& zicp(ji) * pdqsold(ji) / (zikq - peqAcoef(ji)) 


it does not need to put cp_air again during calculating 'pdqsold'

(2) it seems the term

"pb(ji)**kappa"

was used to calculate potential temperature. In orchidee,

(P_r/P)^{kappa}=1

so we do not need to convert the temperature here. and the

"pb(ji)**kappa"

term in 'pdqsold' is not needed.

  1. Proposed revision
pdqsold(ji) = dev_qsol(ji)
  1. Results

By using the revised equation for pdqsold, the oscillations were significantly reduced.

Change History (1)

comment:1 Changed 6 years ago by jgipsl

  • Resolution set to fixed
  • Status changed from new to closed

Done in rev [2093]

Note: See TracTickets for help on using tickets.