cpld_gridgen program does not work on WCOSS2.

[GeorgeGayno-NOAA]

Something is not working. Looks like an ESMF error.

[GeorgeGayno-NOAA]

I just ran the cpld_gridgen regression test on Cactus using 'develop' at ca9bed8. It ran to completion. I don't know what changed on WCOSS2 since I opened this issue. @DeniseWorthen, can you or someone on your team help me evaluate the results?

[DeniseWorthen]

The tools I generally use to inspect the files are not available on wcoss2. Would you be able to put the files on Hera for me to evaluate?

[GeorgeGayno-NOAA]

The baseline data was moved from Hera to Cactus. Here are the file differences from the Hera baseline:

0.25-degree test

Variable  Group Count          Sum      AbsSum          Min         Max       Range         Mean      StdDev
land_frac /     90300 -7.76871e-12 7.71727e-10 -1.55509e-12 2.08733e-12 3.64242e-12 -8.60322e-17 1.74286e-14
Variable  Group Count         Sum     AbsSum          Min         Max       Range        Mean      StdDev
land_frac /     92755 2.12493e-12 6.3463e-10 -1.01841e-12 9.53015e-13 1.97142e-12 2.29091e-17 1.36747e-14
Variable  Group Count         Sum      AbsSum          Min         Max       Range         Mean      StdDev
land_frac /     72576 -3.7161e-12 5.19475e-10 -1.77858e-13 1.16018e-13 2.93876e-13 -5.12029e-17 1.04768e-14
Variable  Group  Count         Sum      AbsSum         Min         Max       Range         Mean      StdDev
land_frac /     132071 -1.3019e-11 1.02577e-09 -2.0195e-12 7.34524e-13 2.75402e-12 -9.85755e-17 1.72286e-14
Variable  Group  Count          Sum      AbsSum          Min         Max       Range         Mean      StdDev
land_frac /     109459 -4.13377e-12 8.34639e-10 -1.65123e-12 1.18383e-12 2.83507e-12 -3.77655e-17 1.49571e-14
Variable  Group  Count          Sum      AbsSum          Min         Max       Range         Mean      StdDev
land_frac /     127224 -4.77261e-13 9.84517e-10 -3.53051e-13 2.21378e-13 5.74429e-13 -3.75134e-18 1.07783e-14
Variable Group   Count          Sum      AbsSum          Min         Max       Range         Mean      StdDev
area_a   /     1524665  3.32988e-17 3.92867e-14 -2.37169e-19 2.40557e-19 4.77727e-19  2.18401e-23  3.5764e-20
area_b   /      861490  1.18491e-16 3.48869e-14 -3.45589e-19 3.50672e-19 6.96261e-19  1.37542e-22 5.16047e-20
frac_a   /      861607 -8.49742e-12 6.97487e-09 -2.20046e-12  2.3832e-12 4.58367e-12  -9.8623e-18 1.48218e-14
frac_b   /      624385 -2.69899e-11 4.77076e-09  -2.0195e-12 2.08733e-12 4.10683e-12 -4.32263e-17 1.44963e-14
S        /     3241434 -2.68928e-11 7.75689e-09 -2.01483e-12 2.08678e-12 4.10161e-12 -8.29657e-18 6.20992e-15
Variable Group Count         Sum      AbsSum          Min         Max       Range        Mean      StdDev
angle    /        13 1.97325e-16 3.57787e-16 -5.55112e-17 1.11022e-16 1.66533e-16 1.51788e-17 4.06852e-17
Variable Group Count        Sum      AbsSum          Min         Max       Range       Mean      StdDev
anglet   /        10 5.5728e-17 2.89916e-16 -1.11022e-16 1.11022e-16 2.22045e-16 5.5728e-18 5.52313e-17

[GeorgeGayno-NOAA]

0.5-degree test

Variable  Group Count         Sum      AbsSum          Min         Max       Range        Mean      StdDev
land_frac /     22623 8.23097e-13 1.03496e-10 -6.65246e-13 1.83853e-13 8.49099e-13 3.63832e-17 8.85007e-15
Variable  Group Count          Sum      AbsSum          Min         Max       Range         Mean      StdDev
land_frac /     23313 -1.42338e-12 8.93204e-11 -4.40981e-13 2.62901e-13 7.03881e-13 -6.10552e-17 7.98463e-15
Variable  Group Count         Sum      AbsSum          Min         Max       Range        Mean      StdDev
land_frac /     18510 5.03627e-13 6.48901e-11 -4.11893e-14 4.25215e-14 8.37108e-14 2.72084e-17 5.12935e-15
Variable  Group Count         Sum      AbsSum          Min         Max       Range        Mean      StdDev
land_frac /     32931 2.76971e-12 1.41934e-10 -3.03757e-13 4.46976e-13 7.50733e-13 8.41065e-17 8.71972e-15
Variable  Group Count         Sum      AbsSum          Min         Max       Range       Mean      StdDev
land_frac /     27326 1.71044e-12 1.15291e-10 -3.26517e-13 2.54019e-13 5.80536e-13 6.2594e-17 7.81463e-15
Variable  Group Count         Sum      AbsSum          Min        Max       Range       Mean      StdDev
land_frac /     31715 7.53296e-13 1.24314e-10 -6.43929e-14 6.9833e-14 1.34226e-13 2.3752e-17 5.39094e-15
Variable Group  Count         Sum      AbsSum          Min         Max       Range        Mean      StdDev
area_a   /     400289 1.35826e-17 2.03978e-14 -4.20128e-19 4.20128e-19 8.40257e-19 3.39319e-23 6.94282e-20
area_b   /     210354 8.12474e-18 1.78667e-14 -6.84403e-19 6.50521e-19 1.33492e-18 3.86241e-23 1.06665e-19
frac_a   /     237535 3.24762e-12 1.01431e-09 -7.28417e-13 5.10481e-13  1.2389e-12 1.36722e-17 8.40645e-15
frac_b   /     156418  5.1368e-12 6.39246e-10 -6.65246e-13 4.46976e-13 1.11222e-12 3.28402e-17 7.51749e-15
S        /     858157 5.07044e-12 1.02366e-09 -6.59806e-13 4.49418e-13 1.10922e-12 5.90852e-18 3.12237e-15
Variable Group Count         Sum      AbsSum         Min         Max Range        Mean StdDev
angle    /         1 2.77556e-17 2.77556e-17 2.77556e-17 2.77556e-17     0 2.77556e-17      0
Variable Group Count Sum AbsSum Min Max Range     Mean   StdDev
col      /        83 -11     83  -1   1     2 -0.13253 0.997204

[GeorgeGayno-NOAA]

1.0-degree test

Variable  Group Count         Sum      AbsSum          Min         Max       Range        Mean      StdDev
land_frac /      5680 5.56754e-14 1.41329e-11 -1.36002e-13 1.08025e-13 2.44027e-13 9.80201e-18 4.76013e-15
Variable  Group Count         Sum     AbsSum          Min        Max       Range      Mean      StdDev
land_frac /      5889 3.55519e-13 1.2329e-11 -7.22755e-14 1.0536e-13 1.77636e-13 6.037e-17 4.22046e-15
Variable  Group Count          Sum      AbsSum          Min         Max       Range        Mean      StdDev
land_frac /      4795 -1.54524e-13 8.46134e-12 -1.53211e-14 1.75415e-14 3.28626e-14 -3.2226e-17 2.56522e-15
Variable  Group Count         Sum      AbsSum          Min         Max       Range       Mean      StdDev
land_frac /      8133 8.80558e-14 1.89385e-11 -9.35918e-14 1.34226e-13 2.27818e-13 1.0827e-17 4.50742e-15
Variable  Group Count         Sum      AbsSum          Min         Max       Range       Mean      StdDev
land_frac /      6829 1.45027e-13 1.47667e-11 -7.39409e-14 8.29337e-14 1.56875e-13 2.1237e-17 3.92277e-15
Variable  Group Count         Sum      AbsSum          Min        Max       Range       Mean      StdDev
land_frac /      7875 5.54109e-13 1.59147e-11 -1.54321e-14 2.4647e-14 4.00791e-14 7.0363e-17 2.75835e-15
Variable Group  Count          Sum      AbsSum          Min         Max       Range         Mean      StdDev
area_a   /     107957 -1.96603e-16 1.13336e-14 -9.21572e-19 9.21572e-19 1.84314e-18 -1.82113e-21 1.44012e-19
area_b   /      50432 -1.67509e-17 9.67612e-15 -1.27394e-18 1.13841e-18 2.41235e-18 -3.32149e-22 2.35753e-19
frac_a   /      63534  3.31224e-12 1.49069e-10 -2.53575e-13 2.86993e-13 5.40568e-13  5.21333e-17 5.18187e-15
frac_b   /      39201  1.04386e-12  8.4543e-11 -1.36002e-13 1.34226e-13 2.70228e-13  2.66285e-17 3.89709e-15
S        /     223246  1.02333e-12 1.34134e-10 -1.32783e-13 1.34559e-13 2.67342e-13  4.58386e-18 1.57841e-15
Variable Group Count Sum AbsSum Min Max Range      Mean StdDev
col      /        15   1     15  -1   1     2 0.0666667 1.0328

[GeorgeGayno-NOAA]

The tools I generally use to inspect the files are not available on wcoss2. Would you be able to put the files on Hera for me to evaluate?

Yes, I can move the files to hera.

[GeorgeGayno-NOAA]

@DeniseWorthen Find the files on Hera: /scratch2/NCEPDEV/stmp1/George.Gayno/CPLD_GRIDGEN.wcoss2.4.denise

[DeniseWorthen]

@GeorgeGayno-NOAA The land_frac differences (these are mapped ocean fractions on the atm tiles) look fine (I spot checked multiple files). The differences are not systematic.

I don't quite understand the report for the variable "col" --it should be a positive integer. It is use in mapping the ocean land mask (src_field) to the ATM tile (dst_field).

    dst_field = 0.0
    do i = 1,n_s
       ii = row(i); jj = col(i)
       dst_field(ii) = dst_field(ii) + S(i)*real(src_field(jj),dbl_kind)
    enddo

I don't understand why the 050 case should report the min as -1? Am I misunderstanding what it is reporting?

[GeorgeGayno-NOAA]

@GeorgeGayno-NOAA The land_frac differences (these are mapped ocean fractions on the atm tiles) look fine (I spot checked multiple files). The differences are not systematic.

I don't quite understand the report for the variable "col" --it should be a positive integer. It is use in mapping the ocean land mask (src_field) to the ATM tile (dst_field).

    dst_field = 0.0
    do i = 1,n_s
       ii = row(i); jj = col(i)
       dst_field(ii) = dst_field(ii) + S(i)*real(src_field(jj),dbl_kind)
    enddo

I don't understand why the 050 case should report the min as -1? Am I misunderstanding what it is reporting?

The report lists the differences between the baseline file (brought over from Hera) and the WCOSS2 computed file. That is why 'col' can be negative. The number of 'col' points that are different is a very small percentage of the total. For the 0.5-degree case, the number of points different is 83 and the total number of points is 867,529.

[GeorgeGayno-NOAA]

@GeorgeGayno-NOAA The land_frac differences (these are mapped ocean fractions on the atm tiles) look fine (I spot checked multiple files). The differences are not systematic.
I don't quite understand the report for the variable "col" --it should be a positive integer. It is use in mapping the ocean land mask (src_field) to the ATM tile (dst_field).

    dst_field = 0.0
    do i = 1,n_s
       ii = row(i); jj = col(i)
       dst_field(ii) = dst_field(ii) + S(i)*real(src_field(jj),dbl_kind)
    enddo

I don't understand why the 050 case should report the min as -1? Am I misunderstanding what it is reporting?

The report lists the differences between the baseline file (brought over from Hera) and the WCOSS2 computed file. That is why 'col' can be negative. The number of 'col' points that are different is a very small percentage of the total. For the 0.5-degree case, the number of points different is 83 and the total number of points is 867,529.

@DeniseWorthen Does my explanation make sense? If so, I would like to close this issue. And turn on the daily regression testing on WCOSS2.

[DeniseWorthen]

Yes, it does. I think you can close the issue. Thanks.

[GeorgeGayno-NOAA]

Sometime after this issue was opened, the cpld_gridgen program started working on WCOSS2. I can't explain it. Will reopen this issue if the problem returns.