the sum of cloud type does not equal to total cloud

[xiaoxiaokuishu]

Dear all,

Recently I checked the value of ("CLDLOW","CLDMED","CLDHGH","CLDTOT") and CLOUD, but the sum of first 3 cloud type does not equal to
CLDTOT, besides CLDTOT and CLOUD are also different . If i plan to use 3-types of cloud and also total cloud, i use first 3 and sum them up as total cloud cover, or use CLDTOT/CLOUD as cloud cover?

Do you have any idea of it.


Best
Ru

 

[brianpm]

Each of those variables (CLDLOW, CLDMED, CLDHGH, CLDTOT) are derived using specific cloud overlap assumptions. They are not additive. They are calculated instantaneously within the model from the CLOUD field, but once a bunch of times are averaged, there's not a good way to derive the 2-dimensional fields from the average 3-d field. See also this recent post: Questions about the cloud variable (eg. CLDLOW)

 

[xiaoxiaokuishu]

Each of those variables (CLDLOW, CLDMED, CLDHGH, CLDTOT) are derived using specific cloud overlap assumptions. They are not additive. They are calculated instantaneously within the model from the CLOUD field, but once a bunch of times are averaged, there's not a good way to derive the 2-dimensional fields from the average 3-d field. See also this recent post: Questions about the cloud variable (eg. CLDLOW)

Thanks for the reply! If i use the total cloud cover (the variable close to MODIS total cloud cover), CLDTOT is the apporiate variable right?

 

[brianpm]

Yes, CLDTOT is the estimate of the total cloud cover, similar to what satellite products produce. Note, however, that the definition of "cloud" in the model is distinct from what satellite retrievals obtain because of assumptions, measurement sensitivity, and observational uncertainty/sampling. A better approach for evaluating cloud cover in a model is to use a satellite simulator to make an artificial retrieval within the model that makes at least the same assumptions as the satellite. See Kay et al. 2012, Medeiros et al. 2023 as examples.