The spatial resolution of local agrometeorological service for farming activities often exceeds the current scales of weather and climate information. When crop models are applied to regional scales, the same problem occurs especially in complex terrain where the coarse meteorological inputs can never represent the underlying land attributes. To supplement the insufficient spatial resolution of official weather and climate information, we suggest a catchment-specific correction scheme based on recently developed techniques in geospatial climatology. Given a single value for a catchment from the nearest weather office, daily minimum temperature at a 30m resolution across the catchment can be produced by the cold air drainage and thermal belt effect embedded in this scheme. Daily maximum temperature can be estimated at the same resolution by applying the solar irradiance effect on sloping surfaces. Rainfall distribution at a 270m grid spacing can be produced by combining the micro-topography effect with the weather RADAR products. We applied this scheme to a mountainous 50km² catchment in South Korea to produce gridded surfaces for temperature, solar radiation, and precipitation at hourly to daily time intervals. Observations at 14 sites showed the validity of this scheme with an operational accuracy. The fine resolution weather data were used to run diverse agronomic models in a site-specific manner together with the localized land attributes such as soils and crop species. Final products in a form of agrometeorological warnings with brief countermeasures were delivered to volunteer farmers to help protect their crops from high impact weather. The economic return from this location-based agrometeorlogical service was compared with that of conventional services using uncorrected weather data and related products.