Simultaneous estimation of daily solar radiation and humidity from observed temperature and precipitation: An application over complex terrain in Austria

Using daily observations of temperature, precipitation, radiation, and humidity from 24 stations spanning a large elevation gradient in Austria, we tested several previously defined algorithms for estimating daily radiation and humidity. The estimation algorithms were first tested independently, and then combined, resulting in a combined algorithm for estimating both radiation and humidity that relies only on temperature and precipitation inputs. Mean absolute errors (MAE) for joint radiation and humidity estimates were 2.52 MJ m^-2 per day and 85.6 Pa, respectively, close to values reported for the algorithm development studies. Biases were low: +0.02 MJ m^-2 per day and +28.2 Pa for radiation and humidity, respectively. Initial results showed biases in estimated radiation related to horizon obstruction and snowpack. We amended the original algorithm, successfully eliminating these effects. Annual prediction MAE was weakly correlated with elevation, and annual bias was not correlated with elevation. Analysis of seasonal patterns in error-elevation relationships showed several periods with significant trends. Radiation MAE was slightly higher in mid-summer for higher elevations, and radiation biases were in general closer to zero throughout the spring and summer at higher elevations. Humidity estimates showed an increased MAE and positive bias at higher elevations in winter. We concluded that the effect of different temperature lapse rates for daily maximum and minimum temperature on the relationship between diurnal temperature range and atmospheric transmittance does not seriously impair predictions over steep elevation gradients in complex terrain. (C) 2000 Elsevier Science B.V.