I changed the solar constant and got the opposite of what what expected.
Original code in radconstants.f90 (/rrtng):
! Solar irradiance at 1 A.U. in W/m^2 assumed by radiation code
! Rescaled so that sum is precisely 1368.22 and fractional amounts sum to 1.0
real(r8), parameter :: solar_ref_band_irradiance(nbndsw) = &
(/ &
12.11_r8, 20.3600000000001_r8, 23.73_r8, &
22.43_r8, 55.63_r8, 102.93_r8, 24.29_r8, &
345.74_r8, 218.19_r8, 347.20_r8, &
129.49_r8, 50.15_r8, 3.08_r8, 12.89_r8 &
/)
My code has half the solar constant:
! Solar irradiance at 1 A.U. in W/m^2 assumed by radiation code
! Rescaled so that sum is precisely 1368.22/2 *** and fractional amounts sum to 1.0
real(r8), parameter :: solar_ref_band_irradiance(nbndsw) = &
(/ &
6.05_r8, 10.1800000000001_r8, 11.85_r8, &
11.22_r8, 27.81_r8, 51.47_r8, 12.15_r8, &
172.9_r8, 109.10_r8, 173.60_r8, &
64.75_r8, 25.08_r8, 1.54_r8, 6.45_r8 &
/)
The result that I got from SCAM6 ARM97 was that temperatures in the bottom level increased by about 10C. it took 48 hours to converge to the new temperatures.
These were my only code changes in that run. I ran the tutorial iop with only my new f90 code in the mods directory. See attached files. (The NetCDF for temperatures is too large to post.)
FYI. I don't intend to modify the solar constant, but I was trying to figure out why I was getting odd results for the changes I made in a different run.
Original code in radconstants.f90 (/rrtng):
! Solar irradiance at 1 A.U. in W/m^2 assumed by radiation code
! Rescaled so that sum is precisely 1368.22 and fractional amounts sum to 1.0
real(r8), parameter :: solar_ref_band_irradiance(nbndsw) = &
(/ &
12.11_r8, 20.3600000000001_r8, 23.73_r8, &
22.43_r8, 55.63_r8, 102.93_r8, 24.29_r8, &
345.74_r8, 218.19_r8, 347.20_r8, &
129.49_r8, 50.15_r8, 3.08_r8, 12.89_r8 &
/)
My code has half the solar constant:
! Solar irradiance at 1 A.U. in W/m^2 assumed by radiation code
! Rescaled so that sum is precisely 1368.22/2 *** and fractional amounts sum to 1.0
real(r8), parameter :: solar_ref_band_irradiance(nbndsw) = &
(/ &
6.05_r8, 10.1800000000001_r8, 11.85_r8, &
11.22_r8, 27.81_r8, 51.47_r8, 12.15_r8, &
172.9_r8, 109.10_r8, 173.60_r8, &
64.75_r8, 25.08_r8, 1.54_r8, 6.45_r8 &
/)
The result that I got from SCAM6 ARM97 was that temperatures in the bottom level increased by about 10C. it took 48 hours to converge to the new temperatures.
These were my only code changes in that run. I ran the tutorial iop with only my new f90 code in the mods directory. See attached files. (The NetCDF for temperatures is too large to post.)
FYI. I don't intend to modify the solar constant, but I was trying to figure out why I was getting odd results for the changes I made in a different run.