APEXv1501 Water Balance Error - SSF Module Issue

[Maya Struzak]

Hello,

I'm experiencing a consistent 5.4% water balance closure error that's unresponsive to parameter changes. After extensive testing and debugging, I suspect a potential groundwater accounting issue.

I am using APEX1501 ver 20250124 to simulate a single subarea as a 5-ha corn field from 1990-1999 following a 10-year spin-up period. In my post-processing code, I have  calculated the total water balance error with daily output values from the .SAD file based on the following equation:

precipitation = WYLD + ET + DPRK + ΔSNO + ΔSW + ΔGWST + ΔSWLT + error

As I mentioned, despite multiple tests adjusting variables related to soil, operation schedule, subarea, as well as controls (i.e. GWST, RFTT, RFPK, SATC, SATK...), I continue to get a 5.4% error across all scenarios tested (over 500mm of water unaccounted for). The .ASA file shows that groundwater is being depleted (-228mm over 10) and the error persists even when groundwater is disabled (GWST=0, RFTT=0). 

The most concerning finding is that substantial RSSF (~400mm/year) is generated even when groundwater storage is set to zero (GWSO=0), suggesting the subsurface flow module may not be properly respecting the groundwater parameter settings. I have attached my .ASA, .OUT, .SUB, .CONT, .SOL, and .OPS files for my simulation with groundwater enabled, as well as a summary of my water balance components for my simulation period (1990-1999). 

1. How can I account for this missing 5.4% of my water balance?

2. Is this a known APEX limitation?

3. Can subsurface flow be completely disabled?

4. Is a 5.4% error acceptable for this system type?

If anyone has encountered any similar issues or has any ideas on other variables to test, I would greatly appreciate any insight!

Thanks in advance.

Best,

Maya

[Jaehak]

Hi Maya,

According to your .OUT file, the average aquifer recharge (PRK) was 443mm, and the GW return flow (RSSF) was 385mm, which seems reasonable. 

1. The percent error on the water balance was 1.77% in the OUT file (line 5657).

2. No, I wasn't aware of this as a limitation.

3. No, subsurface flow can not be turned off or on. What you can experiment with would be that you set soil hydraulic conductivity to zero and GWSO=0, then there will be no PRK and therefore zero RSSF would be expected. 

4. I would say less than 3% error is surely acceptable. >5% error would be the case I would check the water balance settings of the model.

Jaehak