Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

CMIP6 CAM6-CESM2 piControl run TOA radiative imbalance output (why rtmt != rsdt - rlut - rsut, and FSNTOA != SOLIN - FSUTOA)

S

simchan_yook@colostate_edu

Member
Hi all,

I found an error in CMIP6 TOA radiative flux output and have a question.
1. I think rlut in CMIP6 CESM2 picontrol value should be fixed.
Currently rtmt is not equal to rsdt - rlut - rsut, because FSNTOA is not equal to SOLIN - FSUTOA.

2. I was wondering where I could get the CESM2 piControl output used for CMIP6 archive, but not for CIMP6 style rsdt,rlut,rsut variables, but for original CAM6 variables such as SOLIN, FSNT, FLNT, FSUTOA?
I've looked at the /glade/collections/cdg/data/CESM2.1_release/b.e21.B1850.f09_g17.CMIP6-piControl.001/atm/proc/tseries/month_1
but it seems like there are only data for simulation period of "110001-120012".

I've found following relationships between CMIP6 standard variable and CAM5 native output from the link below:

WDCC - Data header information EUUWCVamMAA111rlutv160520

cera-www.dkrz.de cera-www.dkrz.de
rtmt = FSNT - FLNT
rlut = FSNTOA - FSNT + FLNT
rsdt = SOLIN
rsut = SOLIN - FSNTOA

Using the above formulations, rtmt should be equal to rsdt - rlut - rsut (since, rsdt − rlut − rsut = SOLIN - SOLIN + FSNTOA - FSNTOA + FSNT - FLNT = rtmt)
And thus, TOA radiative imbalance can be represented as either 1) rsdt-rlut-rsut or 2) rtmt.

So I've compared those between
1) data set from cdg database
/glade/collections/cdg/data/CESM2.1_release/b.e21.B1850.f09_g17.CMIP6-piControl.001/atm/proc/tseries/month_1
2) CMIP6 archive
/glade/collections/cdg/data/CMIP6/CMIP/NCAR/CESM2/piControl/r1i1p1f1/Amon/

However, it seems like rsdt - rlut - rsut doesn't match the rtmt values, and the difference between rtmt and rsdt - rlut - rsut is quite large (~0.6 W/m^2 on average ?)...

I've also found that the following relationships should be held:
FSNT = SOLIN - FSUT, and
FSNTOA = SOLIN - FSUTOA

Thus, FSUTOA and FSNTOA can be rewritten as:
FSUTOA = SOLIN - FSNTOA = rsut
FSNTOA = SOLIN - FSUTOA = rsdt - rsut

However, somehow FSNTOA is slightly different from SOLIN - FSUTOA (thus also rsdt - rsut), which results in
rsut = SOLIN - FSNTOA != FSUTOA
(I don't know why this is the case...)

Currently I believe following formulations are used for calculating rtmt, rlut, rsut, rsdt:
rtmt = FSNT - FLNT
rlut = FSNTOA - FSNT + FLNT
rsdt = SOLIN
rsut = FSUTOA

However, since FSNTOA != SOLIN - FSUTOA,
rsdt -rlut - rsut = SOLIN - FSNTOA + FSNT - FLNT - FSUTOA is also not equal to FSNT - FLNT

I think rlut should be re-calculated as SOLIN - FSUTOA - FSNT + FLNT rather than FSNTOA - FSNT + FLNT to close the TOA radiative budget
(or FSNTOA = SOLIN - FSUTOA should be hold...)

I've attached some timeseries for those variables below.
rtmt
[138.19968 113.89084 71.45009 -4.42807 -41.41887 -66.68628 -76.9758 -37.713562 29.69632 62.660034]
FSNT - FLNT
[138.19968 113.89084 71.45009 -4.42807 -41.41887 -66.68628 -76.9758 -37.713562 29.69632 62.660034]

rlut
[236.14816 237.1524 251.73232 253.33165 246.45473 268.34344 257.3753 264.72858 284.75198 279.18527]
FSNTOA - FSNT + FLNT
[236.14816 237.15244 251.7323 253.33165 246.45473 268.34348 257.3753 264.72855 284.75195 279.18524]
SOLIN - FSUTOA - FSNT + FLNT
[236.14745 237.15167 251.73169 253.33112 246.45428 268.34305 257.37488 264.7281 284.75146 279.18466]

rsdt
[486.4071 460.97177 412.51703 350.41965 294.83875 266.0003 277.11597 322.73325 382.5498 438.44073]
SOLIN
[486.4071 460.97177 412.51703 350.41965 294.83875 266.0003 277.11597 322.73325 382.5498 438.44073]

rsut
[112.05998 109.92922 89.335266 101.516594 89.80334 64.343506 96.71685 95.71871 68.102066 96.59605 ]
FSUTOA
[112.05998 109.92922 89.335266 101.516594 89.80334 64.343506 96.71685 95.71871 68.102066 96.59605 ]


Thank you,
Shim
 
S

strandwg

Moderator
Staff member
simchan_yook@colostate_edu said:
Hi all,

I found an error in CMIP6 TOA radiative flux output and have a question.
1. I think rlut in CMIP6 CESM2 picontrol value should be fixed.
Currently rtmt is not equal to rsdt - rlut - rsut, because FSNTOA is not equal to SOLIN - FSUTOA.

2. I was wondering where I could get the CESM2 piControl output used for CMIP6 archive, but not for CIMP6 style rsdt,rlut,rsut variables, but for original CAM6 variables such as SOLIN, FSNT, FLNT, FSUTOA?
I've looked at the /glade/collections/cdg/data/CESM2.1_release/b.e21.B1850.f09_g17.CMIP6-piControl.001/atm/proc/tseries/month_1
but it seems like there are only data for simulation period of "110001-120012".

I've found following relationships between CMIP6 standard variable and CAM5 native output from the link below:

WDCC - Data header information EUUWCVamMAA111rlutv160520

cera-www.dkrz.de cera-www.dkrz.de
rtmt = FSNT - FLNT
rlut = FSNTOA - FSNT + FLNT
rsdt = SOLIN
rsut = SOLIN - FSNTOA

Using the above formulations, rtmt should be equal to rsdt - rlut - rsut (since, rsdt − rlut − rsut = SOLIN - SOLIN + FSNTOA - FSNTOA + FSNT - FLNT = rtmt)
And thus, TOA radiative imbalance can be represented as either 1) rsdt-rlut-rsut or 2) rtmt.

So I've compared those between
1) data set from cdg database
/glade/collections/cdg/data/CESM2.1_release/b.e21.B1850.f09_g17.CMIP6-piControl.001/atm/proc/tseries/month_1
2) CMIP6 archive
/glade/collections/cdg/data/CMIP6/CMIP/NCAR/CESM2/piControl/r1i1p1f1/Amon/

However, it seems like rsdt - rlut - rsut doesn't match the rtmt values, and the difference between rtmt and rsdt - rlut - rsut is quite large (~0.6 W/m^2 on average ?)...

I've also found that the following relationships should be held:
FSNT = SOLIN - FSUT, and
FSNTOA = SOLIN - FSUTOA

Thus, FSUTOA and FSNTOA can be rewritten as:
FSUTOA = SOLIN - FSNTOA = rsut
FSNTOA = SOLIN - FSUTOA = rsdt - rsut

However, somehow FSNTOA is slightly different from SOLIN - FSUTOA (thus also rsdt - rsut), which results in
rsut = SOLIN - FSNTOA != FSUTOA
(I don't know why this is the case...)

Currently I believe following formulations are used for calculating rtmt, rlut, rsut, rsdt:
rtmt = FSNT - FLNT
rlut = FSNTOA - FSNT + FLNT
rsdt = SOLIN
rsut = FSUTOA

However, since FSNTOA != SOLIN - FSUTOA,
rsdt -rlut - rsut = SOLIN - FSNTOA + FSNT - FLNT - FSUTOA is also not equal to FSNT - FLNT

I think rlut should be re-calculated as SOLIN - FSUTOA - FSNT + FLNT rather than FSNTOA - FSNT + FLNT to close the TOA radiative budget
(or FSNTOA = SOLIN - FSUTOA should be hold...)

I've attached some timeseries for those variables below.
rtmt
[138.19968 113.89084 71.45009 -4.42807 -41.41887 -66.68628 -76.9758 -37.713562 29.69632 62.660034]
FSNT - FLNT
[138.19968 113.89084 71.45009 -4.42807 -41.41887 -66.68628 -76.9758 -37.713562 29.69632 62.660034]

rlut
[236.14816 237.1524 251.73232 253.33165 246.45473 268.34344 257.3753 264.72858 284.75198 279.18527]
FSNTOA - FSNT + FLNT
[236.14816 237.15244 251.7323 253.33165 246.45473 268.34348 257.3753 264.72855 284.75195 279.18524]
SOLIN - FSUTOA - FSNT + FLNT
[236.14745 237.15167 251.73169 253.33112 246.45428 268.34305 257.37488 264.7281 284.75146 279.18466]

rsdt
[486.4071 460.97177 412.51703 350.41965 294.83875 266.0003 277.11597 322.73325 382.5498 438.44073]
SOLIN
[486.4071 460.97177 412.51703 350.41965 294.83875 266.0003 277.11597 322.73325 382.5498 438.44073]

rsut
[112.05998 109.92922 89.335266 101.516594 89.80334 64.343506 96.71685 95.71871 68.102066 96.59605 ]
FSUTOA
[112.05998 109.92922 89.335266 101.516594 89.80334 64.343506 96.71685 95.71871 68.102066 96.59605 ]


Thank you,
Shim
Click to expand...

The data header you point to is from data submitted to CMIP5 by a group at the University of Washington back in 2015, so the original CESM output aren't necessarily directly comparable to what was submitted to CMIP6, as CAM changed in the interim.

For CMIP6, the CAM6 equivalents for rtmt, rsdt, rlut and rsut are as follows:

rlut = FLUT
rsdt = SOLIN
rsut = FSUTOA
rtmt = FSNT - FLNT

For CMIP5, the CAM5 equivalents were

rlut = FSNTOA-FSNT+FLNT
rsdt = SOLIN
rsut = SOLIN - FSNTOA
rtmt = FSNT - FLNT
 
S

simchan_yook@colostate_edu

Member
strandwg said:
The data header you point to is from data submitted to CMIP5 by a group at the University of Washington back in 2015, so the original CESM output aren't necessarily directly comparable to what was submitted to CMIP6, as CAM changed in the interim.

For CMIP6, the CAM6 equivalents for rtmt, rsdt, rlut and rsut are as follows:

rlut = FLUT
rsdt = SOLIN
rsut = FSUTOA
rtmt = FSNT - FLNT

For CMIP5, the CAM5 equivalents were

rlut = FSNTOA-FSNT+FLNT
rsdt = SOLIN
rsut = SOLIN - FSNTOA
rtmt = FSNT - FLNT
Click to expand...

Thank you for your clarification!

I've checked that the last 500 years of rtmt has very small net mean value (0.028 W/m^2; consistent with Danabasoglu et al., 2020), but the "rsdt-rsut-rlut" has a global mean value of 0.635 W/m^2.

I was wondering whether the new definition of CMIP6 variables for the net TOA radiative fluxes (rsdt-rsut-rlut) resulting in a net positive global-mean values (0.64 W/m^2 in global average) is intended to differentiate the net radiative fluxes at TOA and those values at the model top? Or did I calculate something wrong?

Thanks again,
Shim
 
S

strandwg

Moderator
Staff member
simchan_yook@colostate_edu said:
Thank you for your clarification!

I've checked that the last 500 years of rtmt has very small net mean value (0.028 W/m^2; consistent with Danabasoglu et al., 2020), but the "rsdt-rsut-rlut" has a global mean value of 0.635 W/m^2.

I was wondering whether the new definition of CMIP6 variables for the net TOA radiative fluxes (rsdt-rsut-rlut) resulting in a net positive global-mean values (0.64 W/m^2 in global average) is intended to differentiate the net radiative fluxes at TOA and those values at the model top? Or did I calculate something wrong?

Thanks again,
Shim
Click to expand...

I'm not expert on radiative fluxes in CAM - someone who is will have to chime in.
 
B

brianpm

Active Member
Yes, I'm pretty sure that you are correct. The "residual at Top of Mode" is: RESTOM = SOLIN - FSUT - FLUT, and this would be the value used for Danabasoglu et al.

The inconsistency comes because rsut = FSUTOA which does not include the "extra layer" used in the radiation scheme.

This is properly defined in terms of the CMIP request:
rtmt = "net_downward_radiative_flux_at_top_of_atmosphere_model"
rsut = "toa_outgoing_shortwave_flux"
rsdt = "toa_incoming_shortwave_flux"
rlut = "toa_outgoing_longwave_flux"

Presumably rtmt has been added precisely because a number of models have this inconsistency/ambiguity between "top of model" and "top of atmosphere".
 
You must log in or register to reply here.
Top