Bug found during VIC science test when comparing VIC4.2.d results with VIC5.0.1
Bug description:
Even with hourly met forcing input of all the 7 required variables (so in theory no need for any met disaggregation), VIC4.2.d sometimes introduces a 1-day shift in all met variables, resulting in a 1-day shift in subsequent simulation results (i.e., if try to output met variables, such as OUT_PREC or OUT_AIR_TEMP, these output met variables are shifted by 1 day compared to input forcing; other output variables are also shifted consequently).
This only happens at some locations but not all (e.g., in the VIC science test suite, 40.0329_-105.5464 has 1-day shift, while 41.1649_-96.4701 has correct timing)
By simply editting the latlon values for a shifted location fixed the issue, so the bug is likely due to some messing up with latlon when processing timestamps; but further investigation and bug fixing is needed
Bug Reports
OUT_PRECorOUT_AIR_TEMP, these output met variables are shifted by 1 day compared to input forcing; other output variables are also shifted consequently).40.0329_-105.5464has 1-day shift, while41.1649_-96.4701has correct timing)latlonvalues for a shifted location fixed the issue, so the bug is likely due to some messing up withlatlonwhen processing timestamps; but further investigation and bug fixing is needed