swatplus-editor@google.com

[ABEBE HEGANO]

SWAT+ QWflow Issue Report  for Assistance

Hello,

SWAT+ QWflow Issue Report

1. Primary Issue Location

• Component: SWAT+ Executable / Model Functionality (gwflow module)
• Area of impact: Model output (simulated streamflow substantially overestimated vs. observed)

2. Software Versions

• GIS Interface: QSWAT+ 4.0.3
• SWAT+ Editor: 4.0.2
• SWAT+ Revision: rev62

3. Project Characteristics

• Watershed Details: Enkare-Narok Basin in Kenya

o   Short channel merge threshold 5%

o   Using percentage of subbasin as a threshold

o   Multiple HRUs Landuse/Soil/Slope option       

o   Thresholds: 5/10/10 [%]

o   Number of subbasins: 30

o   Number of channels: 134

o   Number of LSUs: 268

o   Number of HRUs: 2766

-   GWgrid cell size (200)

• Modeling Aim: Water balance / streamflow, sediment and crop yield  
• gwflow was adopted specifically to address a transmission loss issue in stream-aquifer exchange prior to reaching the outlet not a groundwater-specific study objective.

4. Description of Issue

The Problem:
Simulated streamflow is persistently and substantially overestimated after enabling gwflow. Observed mean flow is 4.364 m³/s. After correcting an aquifer/HRU thickness input error, simulated mean flow improved from 41.4 to 31.97 m³/s (PBIAS 848.8% → 632.7%; NSE -113.685 → -74.945), but remains roughly 7.3x observed. Prior to gwflow (classic groundwater module), PBIAS on the same basin was approximately +118%, indicating gwflow introduced a large additional inflation beyond pre-existing bias.

I've seen a related unresolved thread on this list "gwflow baseflow discharge (gw_sw) not responding to recharge or aquifer parameters “SWAT+ rev. 62" and am posting in case my symptoms are related, or informative for that diagnosis.

v  Hydraulic conductivity (GLHYMPS-derived, ~0.273 m/day bedrock, up to 20+ m/day alluvium) appears within expected range for these units.

v  gwflow is intended to fully replace the classic aquifer module; have not yet independently confirmed the classic aquifer's GW_Q term is zero in basin output.

v  The CONST column in gwflow.con is zero across all rows — unclear whether this is expected default behavior or an unpopulated streambed conductance value.

Happy to share TxtInOut files and looking for your assistance

https://drive.google.com/drive/folders/1-2H9FgLhRU90sJ0UuB4ncSvkyf2KCws4?usp=drive_link