Extrapolation of Daily Maximum Temperature in a Mountainous Terrain

Since temperature at canopy-level is a critical determinant of many biological processes, extrapolation of synoptic data is common to get site-specific temperature data for the purpose of agricultural and natural resources management. Besides elevation, exposure of the site is believed to play an important role in determination of daily maximum temperature in complex terrain. We suggest a simple method using a normalized difference in incident solar energy between a target surface and a level surface (NDSI) to eliminate errors in daily maximum temperature extrapolation in mountainous complex terrain. We measured air temperature at 8 side slope locations circumventing a cone-shaped parasitic volcano (c.a., 570 m diameter for the bottom circle and 90 m bottom-to-top height) covered with natural grasses in Jeju Island, South Korea. A dataset of daily maximum temperature was prepared from summer to winter solstice in 2007 and deviations at each location from a horizontal reference were regressed to the NDSI, to search a quantitative relationship between daily maximum temperature and solar irradiance. The selected nonlinear regression equation (0.76 for the coefficient of determination) was used to extrapolate daily maximum temperature over a mountainous watershed in 2008. When the NDSI corrected temperatures at selected sites were compared with the best official estimates from national automated weather observation network (NAWON), a consistent reduction in the root mean square error (RMSE) was observed for the entire period. Seasonal averages for the RMSE were 0.83 ℃ for the NDSI corrected estimation compared with 1.50 ℃ for the NAWON based interpolation.