I have a .csv file containing data for precipitation, discharge, and increase in stream discharge over 15 minute intervals from 2010 until 2022. The format of my file is
Date, Time, Discharge, IncreaseDischarge, Precipitation
2010-05-01, 0:00:00, 0.299, 0.0, 0.0
2010-05-01, 0:15:00, 0.319, 0.020, 4.6
2010-05-01, 0:30:00, 0.371, 0.052, 3.1
In the above .csv file, "Date" is the date ranging from 2010-05-01 to 2022-10-31, "Time" is the time of day, "Discharge" is the discharge measurement from the hydrometric streamflow gauge at each time interval, "IncreaseDischarge" is the increase in stream discharge at each time interval (compared to the discharge at the previous time interval), and "Precipitation" is the precipitation depth that has fallen over each time interval.
Incremental disturbances have been added to the watershed (2013, 2016, and 2019), and I want to be able to test whether (and the extent to which) these disturbances change the "IncreaseDischarge" in response to "Precipitation". i.e., is the relationship between precipitation depth and the corresponding increase in stream discharge statistically different before 2013, between 2013 and 2016, and after 2019? Does anyone have any advice on how to approach this? I have no idea where to start, I started trying to create GAM models for the relationship between streamflow and rainfall depth, but the GAM models did not consider rainfall intensity, which has a large role on the streamflow response.
I would appreciate any feedback or suggestions that anyone may have as I have no idea what I am doing at this point. TIA