Questions about the cloud variable (eg. CLDLOW)

[Yiqun]

Hello, I have two questions about the cloud output:

1. How does the Vertically-integrated low cloud (CLDLOW) is calculated? Like what's the pressure level range for low/mid/high cloud and if it's just the simple average between different levels or weighted average or else?

2. There are three variables about low cloud: a
a. Lidar Low-level Cloud Fraction (CLDLOW_CAL)
b. MODIS Low Level Cloud Fraction (CLLMODIS)
c. Vertically-integrated low cloud (CLDLOW)
What's the difference of these three variable about low cloud? I took a look at them and they looks different.

It would be very helpful if you have any comments about these two questions or could guide me to some related documents. Thank you!

 

[brianpm]

For #1:
CLDLOW is calculated in the cldsav subroutine of cloud_cover_diags.F90. It is calculated using a maximum overlap assumption in the region from the surface to 700hPa. Think of it as a projection of the cloud cover onto the surface rather than an "average."

For #2:
The CLDLOW output uses the model's native definition of cloud fraction. The other use definitions that are more comparable to satellite observations by using COSP (COSP: Satellite simulation software for model assessment, GMD - The Cloud Feedback Model Intercomparison Project Observational Simulator Package: Version 2). The lidar cloud fraction is comparable to the CALIPSO retrieval, and the MODIS one is comparable to MODIS. There are similar cloud fractions for MISR and ISCCP. Examples of how to use these are provided by Kay et al. (2011, Exposing Global Cloud Biases in the Community Atmosphere Model (CAM) Using Satellite Observations and Their Corresponding Instrument Simulators) and Medeiros et al. (2023, https://doi.org/10.1029/2023EA002918)

 

[Yiqun]

Thank you so much for your reply! Could you please upload the file cloud_cover_diags.F90 again? I can't open it. Thank you!

 

[brianpm]

Sorry about that, it's just a github link, but I think my copy didn't work correctly:

CAM/src/physics/cam/cloud_cover_diags.F90 at cam_cesm2_2_rel · ESCOMP/CAM

Community Atmosphere Model. Contribute to ESCOMP/CAM development by creating an account on GitHub.

github.com

 

[xiaoxiaokuishu]

Hi, yiqun,

I did not find the CLLMODIS in the default output from cam, is there some additional flag should be set to get such a varibale.

Thanks

 

[brianpm]

That is correct. CLLMODIS is derived from CLMODIS (which has dimensions [time, cloud-top-pressure, cloud-optical-thickness, lat, lon]). The simulators are provided by COSP (Swales et al. 2018). To turn on COSP in CAM requires:

xmlchange --append CAM_CONFIG_OPTS="-cosp"

(See user guide: 3. Building and Running the atmospheric model within CESM — camdoc documentation)

You will need to set namelist settings to configure COSP if you want any of the non-default settings. See: CESM2.2.0 CAM6.3 Namelist Definitions (and search for "cosp" using the search field). The output fields are added in the usual way using FINCL (see users guide: 8. Model Output: — camdoc documentation)