get_cloudfrac

[ukmetmodeller]

I've just been looking at the WRF-python code for calculating cloud cover. The python script uses the DCLOUDFRAC2 from wrf_cloud_fracf.f90 which all makes sense. I understand all elements of the process but I believe that the low / mid / high grouping is wrong when using pressure.

low_thresh = 97000
mid_thresh = 80000
high_thresh = 45000

I believe the requirements should be:
Low: full_p(k) <= low_thresh & > mid_thresh
Mid: full_p(k) <= mid_thresh & > high_thres
High: full_p(k) <= high_thresh

See below table for pressure levels at a single location (i,j)

k

1​

2​

3​

4​

5​

6​

7​

8​

9​

10​

11​

12​

13​

14​

15​

16​

17​

18​

19​

20​

21​

full_p

96580.59​

96057.03​

95317.11​

94014.46​

92151.03​

89689.49​

86636.23​

82997.47​

78777.13​

73980.17​

68599.44​

62624.09​

56031.54​

48747.43​

40990.8​

33650.34​

27540.33​

22820.15​

18995.81​

15398.71​

11799.98​

<= low_thresh & >mid_thresh

9.17E-01​

0.951522​

1​

1​

1​

1​

1.00E+00​

7.47E-01​

<= mid_thresh & >high_thresh

0.172427​

0​

0​

0​

0​

0​

<= high_thresh

0.145057​

0​

0​

0​

0​

0​

0​

This would put level k=8 into the low group (82997.47 Pa is greater than 80000 Pa and less than 97000).
However, DCLOUDFRAC2 puts it into the mid group based on the following fortan code:

kchi = -1
kcmi = -1
kclo = -1

DO k = 1,nz
IF (vert(i,j,k) .GT. low_thresh) kclo=k
IF (vert(i,j,k) .GT. mid_thresh) kcmi=k
IF (vert(i,j,k) .GT. high_thresh) kchi=k
END DO

Here kclo=-1, kcmi=8, kchi=14

Then base on the loop below pressure level k=8 (82997.47 Pa) is greater than or equal to kcmi and less than kchi so fits into the mid cloud group.
DO k = 1,nz
IF (k .GE. kclo .AND. k .LT. kcmi) THEN
lowc(i,j) = MAX(rh(i,j,k), lowc(i,j))
ELSE IF (k .GE. kcmi .AND. k .LT. kchi) THEN ! mid cloud
midc(i,j) = MAX(rh(i,j,k), midc(i,j))
ELSE if (k .GE. kchi) THEN ! high cloud
highc(i,j) = MAX(rh(i,j,k), highc(i,j))
END IF
END DO

This doesn't match the criteria of Mid: full_p(k) <= mid_thresh & > high_thres (mid <= 80000 and > 45000)

Does anyone know if this is a mistake or intentional?

 

[William.Hatheway]

 # Get the cloud fraction data
    cfrac = wrf.cloudfrac(
        vert=height,
        relh=rh,
        vert_inc_w_height=1,
        low_thresh="1800",
        mid_thresh="3600",
        high_thresh="5400",
        meta=True,
    )

    # Mid Cloud Fraction
    Mid_cfrac = cfrac[1, :, :] * 100.0

I use WRF-python but this might help

wrf.cloudfrac — wrf-python 1.3.4.1 documentation