I would like to estimate stormwater runoff using a methodology that takes into account both land surface cover and climatic conditions. Is there a reliable approach to achieve that?
Accuracy and flow estimation with just climate and land cover data do not go well together concerning stormflow. The total flow during a storm includes some amount of groundwater baseflow, interflow and runoff. End member analysis has been used to separate these elements with some success, by Garrett et al and others. If you are dealing with small catchments or small streams, some type of flume or v notch weir and water level recording may be best for accuracy. In a wadable stream with the right current meters, one can probably do a respectable estimate with some effort, recognizing that any depth (ft) times velocity (ft/sec) should be less than 10 for safety as a rule of thumb. Without a weir or flume, sometimes a culvert might be used to estimate flow with water level recordings.
Estimating stormwater runoff is not be just estimating the storm peak but estimating the total volume of flow and using techniques, to separate the baseflow from the stormflow. The separation is an estimate.
Other methods of flow estimation are possible with methods such as injecting fluorescent dye at known concentration at a constant rate and using sampling and a fluorometer to record the concentration change with time.
If you had a pond downstream of area of interest, you migh be able to collect the stormflow and measure the volume change in pond. If the channel has bankfull indicators, and you know how to read them, with the proper measurements of channel cross section, you might come up with a good estimate of the bankfull storm peak value.
use NRCS web site. Also, number of softwares are also available / EPS SWMM 1.5/ Hydrognomon etc. search in the google for softwares.
However, SCS curve number method is more popular. more details, read FAO irrigation and drainage working papers and related text books. http://www.fao.org/3/ah832e/ah832e00.pdf