Articles
TEMPERATURE ACCURACY OF AGRICULTURAL WEATHER STATIONS AND CONSEQUENCES FOR PHENOLOGY MODELLING
Article number
803_12
Pages
105 – 110
Language
English
Abstract
The network of agricultural weather stations for local plant alert service operated by the research station ACW and the Swiss cantons consists of appliances from different manufacturers that show obvious differences in temperature measurements.
In order to assess which stations can be used for reliable alert service the present study evaluates the quality of the temperature data of the following devices in comparison to high precision weather stations of the Swiss national weather service (MeteoSchweiz): HP-100, OPUS 200 (Standard), OPUS 200 (Side arm) (all of LUFFT, Germany), and Campbell CR10X (Campbell Scientific (UK).
Especially on high-radiation days all sensors except for that of the Campbell station overshot the high precision measurements in part drastically during daytimes.
Average daily temperatures showed a significant but slight deviation in the OPUS 200 stations whereas they did not differ significantly in Campbell and HP-100, in the latter night lows averaging out the higher daytime values.
Forecasts for disease alert that is based on average daily temperatures or temperature sums are therefore only slightly influenced by the weather station used.
Daily maximum temperatures and average hourly temperatures during daytime differed significantly and considerably from high precision measurements in all station except for the Campbell.
Disease forecasts on base of daily maximum temperatures could therefore be heavily influenced by the weather station in use with only Campbell showing satisfactory results.
Deviations in hourly data were obviously also reflected in the bias of simulated phenology of codling moth (Cydia pomonella) being up to 10 days too early for adult flight and 9 days for larval hatch in the OPUS 200. The only tested agrometeorological station sufficient for pest phenology modelling again was the Campbell CR10X.
In order to assess which stations can be used for reliable alert service the present study evaluates the quality of the temperature data of the following devices in comparison to high precision weather stations of the Swiss national weather service (MeteoSchweiz): HP-100, OPUS 200 (Standard), OPUS 200 (Side arm) (all of LUFFT, Germany), and Campbell CR10X (Campbell Scientific (UK).
Especially on high-radiation days all sensors except for that of the Campbell station overshot the high precision measurements in part drastically during daytimes.
Average daily temperatures showed a significant but slight deviation in the OPUS 200 stations whereas they did not differ significantly in Campbell and HP-100, in the latter night lows averaging out the higher daytime values.
Forecasts for disease alert that is based on average daily temperatures or temperature sums are therefore only slightly influenced by the weather station used.
Daily maximum temperatures and average hourly temperatures during daytime differed significantly and considerably from high precision measurements in all station except for the Campbell.
Disease forecasts on base of daily maximum temperatures could therefore be heavily influenced by the weather station in use with only Campbell showing satisfactory results.
Deviations in hourly data were obviously also reflected in the bias of simulated phenology of codling moth (Cydia pomonella) being up to 10 days too early for adult flight and 9 days for larval hatch in the OPUS 200. The only tested agrometeorological station sufficient for pest phenology modelling again was the Campbell CR10X.
Authors
M. Sacchelli, W. Siegfried, J. Samietz
Keywords
plant protection, weather data, precision, temperature measurements, simulation bias, disease model, pest phenology model
Online Articles (41)