Generating Daily Surfaces of Temperature and Precipitation over Complex Topography

Document Type

Book Chapter

Publication Title

GIS and Environmental Modeling: Progress and Research Issues.

Publication Date



Chapter 18

First Page


Last Page



We present the logical and algorithmic framework of a numerical model which generates daily interpolated surfaces of maximum air temperature, minimum air temperature, and precipitation over a gridded terrain, and we demonstrate its application in the state of Montana for a one-year period. The model generates daily estimates of the relationship between each meteorological variable and elevation for each grid cell in the terrain model and uses these estimates to generate interpolated surfaces of temperature and precipitation based on daily observations from a network of recording stations. Interpolates are based on a vertically exaggerated proximal polygon algorithm, in combination with a Gaussian-weighted spatial convolution kernel. Elevation relationships are estimated with a least squares weighted linear regression model that varies in space and in time by subsetting both the spatial and temporal domains. Standard lapse-rate regressions are used to model the relationship between maximum and minimum temperature and elevation. Precipitation relationships are based on a normalized difference algorithm. Predictions of precipitation occurrence employ a local event frequency statistic. We present some of the principal results of simulations across the state of Montana on a 1-km resolution grid for 1990. Cross-validation was used to generate daily predictions of maximum and minimum temperature and precipitation. Predicted annual averages were compared to observed annual averages, resulting in mean absolute errors for daily prediction of maximum and minimum temperatures of 0.69 degrees and 0.98 degrees Celsius/day, respectively. Mean absolute errors for predicted precipitation were 0.03 cm/day (11.83 cm/year, or 20.0% measured as a proportion of total annual precipitation).


Editors: Goodchild, M., Steyaert Parks Johnston Maidment Crane Glendinning L. B. C.


© 1996 GIS World, Inc.

This document is currently not available here.