The baseflow recession curve technique consists of interpreting the evolution of the flow rate of a river (Q vs. t) to assess the hydraulic interactions between surface water flowing in the river and groundwater (the water table connected to the river). During a recession period the beghaviour of the river flow rate decrease gives an idea of the water that is lost by the river to the aquifer. This part of the river water "lost" in the aquifer is assumed to be an estimate of the recharge of the aquifer during the recession period.

The log behaviour of the decrease of the river flowrate is found when applying a log-regression fit on the curve which then allows to determine the recession constant K. There is no rule to apply this fit which needs to be done by the person interpreting the graph (subjective interpretation).

Well,

can be found in page 177 to 184 from the textbook. introduction to hydrology. by/ W. Viessman

Sincerely

I typically use the midpoint of the slope inflection point that marks the change between the hydrograph recession of storm, and the baseflow recession slope. Extend both slopes and it marks the intersection, find the inflection from that point by 1/2 of the angle up to the hydrograph. This is a physical approach, but some use a chemical or end member approach. Certainly, consult hydrology texts and papers on subject. I will try to return later and attach a paper by Garrett et al. in using end member analysis that is more intensive than physical hydrograph approach. Complexity is added when multiple storms occur in sequence. When there is a substantial difference between rainfall chemistry and groundwater, the chemical approach may be helpful, and you can compare with physical hydrograph approach. Sorry, I cannot comment on the papers listed.

Also please note, the recession slope may vary somewhat with conditions, such as growing season vs dormant season for example. Certain conditions such as droughts and severe wildfires can alter runoff, and limit recharge, altering normal hydrograph responses and recession curves temporarily or short term.

Hi,

As William F. Hansen pointed out, point of inflection is a good starting point. Inspection of the Log of the -dq in flow might reveal other structures. Also I try to exclude some data after the peak (Beck et al, 2013 chose ~5days, but his rivers were larger), this is usually chosen based on the Strahler order of the stream. This offset is to eliminate any Rapid flow patterns after the peak and is a bit trail and error. You would also want to determine what kind of model do you want to fit for the recession storage, a linear or a non-linear. Usually clubbing together all the recession limbs in a time series, it becomes evident that one log(-dq/dt) structure repeats itself for most the limbs. Take avg of values for all limbs or get model coeffcients from clubbed recession data. Also you might want to classify the recession limbs base on wet or dry years. In a wet or flood year the recession limb might be affected by other conditions or might not really represent baseflow recession if another peak is closeby.

You could also look at the papers by Lamb and Beven 1997 and Beven et al 2011 on generating a master recession curve

In general hydrology is a science of prediction and forecasting, I believe choosing any method mentioned in books of (V.T Chow, Subramania, Vector Ponce, Nurunnisa USULm, Ghanshyam Das, and others.....) gives close results and accepted within prediction and forecasting

Anand Nainar thanks for both your suggestion and adding documents.

Well, for the choose of inflection points it's mostly based on automated calibration. however, sometimes this step of analysis can be considered objective, especially when we do that visually based on the ground slope.