Agriculture in hill and mountain ecosystems is predominantly rainfed with common occurrence of moisture stress. It is a natural disaster which evolves in time and its impacts last for a long time. In the present investigation, long-term monthly precipitation data for 40 years (1980–2019) were used for characterizing agricultural drought in Almora and Nainital districts of Uttarakhand in India. Different drought indices based on meteorological data like standard precipitation index (SPI), percentage of departure (P$_{d}$) and percent of normal (P$_{n}$) were used. Percentage of departure is calculated from deviation of monthly precipitation from the long-term average monthly precipitation. Percent of normal is calculated by dividing the precipitation by normal precipitation for time being considered. SPI values were calculated based on gamma distribution of long-term monthly precipitation data. The Pearson’s correlation coefficient between monthly percentage of departure and different SPI time scales (1, 3 and 6 months) were analyzed. SPI-1 (July and August) for both the stations showed very strong correlation with the corresponding monthly percentage of departure (r  > 0.97) than SPI-3 and SPI-6. Therefore, it is suggested that SPI as a stand-alone indicator should not be interpreted to identify drought in a hilly region.

$\bf{Highlights}$

$\bullet$ Meteorological drought indices have been used to identify agricultural drought.

$\bullet$ SPI-1 showed very strong correlation with percentage of departure.

$\bullet$ Meteorological based SPI was well correlated with satellite based drought indices.

$\bullet$Study suggest to use multiple drought indices for drought Identification.