In order to model the discharge of rivers, I use the relationship Qpike =RR*P (RR=1-1/(1+(P/PET)^2)^0.5) : depending on precipitation (P) and potential evapotranspiration (PET). But the modeled values are generally lower than the observed values. My question how to correct and adjust the model in order to have a good correlation
It is unusual that one equation like this can do well at defining discharge in rivers, especially daily or instantaneous. It might do OK for monthly or annual values if developed from your observed river data through time. The approach appears more of a water balance, but might be good to add a factor for soil moisture change. I uploaded a report by Feaster et al and it is based on a grid of stream gauges to predict frequency of various storm flows. Perhaps that is not your intent, maybe your equation is generally fine for mean flow, but needs some tweeking for land use. If your existing equation underestimates by a small amount, try to include more runoff from impervious surfaces, grasslands or cultivated fields as compared to forest lands. I would carry out equation coefficients and exponents to more than one figure and look into how equation was developed, empirical or based on data. But any equation you can come up with that is within plus or minus 20% for your intended needs is probably not that bad.
Thank you William F. Hansen for your answer
I use this model (Qpike) to estimate the average annual values of water discharge (long term)
it's interesting to integrate soil moisture data, but I can not find the data (global map .tif), which present average annual values of soil moisture (long-term)!!!
- Badges
- Science topic
- Similar topics
- Geoscience
- Hydrology