In the case of arid and semi-arid reagion, what is best, either use the normal Curve Number (CN for normal condtion) or adjusted Curve Number (CN for wet and dry conditon) to estaimate the poetintial runoff from rainfall data?
Dry, normal and wet conditions have a strong impact on runoff. There's no simple answer.
Theoretically, you should make an analysis for understanding the link between soil moisture conditions, rainfall and runoff. The selection of one condition arbitrarily might produce strong errors.
I think this depends on hydro-climatic condition of the watershed. If one is able to validate for dry and wet periods then dry and wet CN seems more plausible.
Dear Ahmed Shahadha Muneer, It is hard to formulate a definitive answer. As Prof. Luca Brocca suggested, you should try for yourself. From a theoretical stand-point, normally, antecedent soil moisture conditions always affect to a certain extent runoff generation processes (in terms of amount and procedure). Therefore, accounting for wet/dry conditions matter. In the case of a study we conducted in a typical sahelian landscape, we found that soil moisture conditions at the onset of a rainfall event were quite critical to capture runoff processes . See the paper below:
Article Modelling the water balance on farming practices at plot sca...
However, only a sound analysis in your typical situation can provide a valid answer. Just compare the two approaches and decide for yourself. And always communicate your findings (in this thread), as it might provide useful incentives to the discussion.