n the mountainous region, springs are the main sources of drinking water supply as well as domestic water, therefore these s
of the region. During summer, spring flow reduces significantly and sometimes they dried
hydrology is complex in nature and lack of a database on springs restricts the application of developed hydrological models
critical waters resources as a sustainable source of water in the re
developed to predict the spring flow for two springs namely, Hill Campus and Fakua from Tehri Garhwal district of Uttarakhand
data of precipitation (P), temperature (T), relative humidity (RH) and spring flow (Q) from the year 1999 to the year 2003 were used to
model the spring flow. First, four-year data was used to train the models, whereas, remaining one
models in predicting the springflow. Seven models differ in input parameters (P, T, RH & Q) have been developed and the best fit model was
selected on the basis of capability of the network to converge normalize system error (NSE). Developed models performance wer
by categorize all data samples into six ranges (0
with observed data. A close agreement between observed and ANN predicted spring flow in the testing period for both the sprin
the capability of ANN in predicting the spring flow even in data scare situation. The study also reveals that variability in
results in better accuracy in the testing period