Integrating statistically generated weather data with climate change in EPIC – a case study analysis
Project Title: Integrating statistically generated weather data with climate change in EPIC – a case study analysis
Collaborating GEO-BENE partners: BOKU (lead), KTL, SSCRI, IIASA
Global climate evaluations show at varying initial conditions (e.g. economic growth, development of the world population and the use of renewable energy) different climate scenarios in the next 100 years. The uncertainty of the climate in the coming 20 to 30 years is relatively small, given that the characteristic time span of significant economic changes and innovations is around 30 years. Therefore, we have developed a statistical model based on the weather observations from 1975 to 2006 to generate weather data considering climate change for the next 30 years. The daily weather observation data are from a weather station in the Austrian region Marchfeld.
Daily weather data from Groß Enzersdorf (Marchfeld / Lower Austria) for the years 1975 to 2006. Representative site, soil and crop management data for Marchfeld.
The EPIC model and linear regression including linear variables and seasonal covariates; redistribution of data to get random predictions. Stochastic bio-physical impact analysis.
First, we have investigated the impacts of climate change on the crop yield and soil organic carbon (see following pictures).
Only one trend is shown for the period 1975-2006, because we use observed weather data. For the next 30 years, we have randomly selected months from the past period and generated a new weather by respecting the monthly sequences (e.g. the weather in March of the past period is also the weather of March in the new seeds) and the trends in daily Tmin and Tmax temperatures and monthly rainfall amounts. This process has been repeated to construct 100 random weather seeds.
Only the crop yield trend of summer barley and the trend of soil organic carbon are significant (see table above) in the period 1975-2006. The table also contains the percentage of significant trends for the period 2007-2038. There is only a significant difference between the two periods for ‘soil organic carbon’.
Next steps will focus on the impacts of climate change in the period 2007-2038 on the profitability of different crop production systems and other environmental indicators.