Bias correction and estimation of climate change effects on yields

GOALS

The overall objective of COMPROMISE is to address the above-mentioned challenges. It consists of two parts:

1. Compare several multivariate bias correction methods applied to climate series that will be used as input variables for several impact models (phenology, water balance, ET0, FWI).
2. Synthesise a wide range of studies to estimate, on a global scale, the changes in crop yields resulting from the marginal and combined effects of temperature, CO2 levels and precipitation. In addition, a thorough assessment of the accuracy of indices (simple or aggregated) based on weather and satellite imagery for predicting yield losses is carried out.

The effect of rising temperatures on crops

Climate change impacts crop yields through higher temperatures, changes in precipitation and increased CO2 levels in the atmosphere. Although the potential effects of each of these factors have been examined in a number of separate studies, no recent synthesis has been published to provide quantitative estimates of the effects of climate change on crop yields, with or without adaptation strategies.

However, a large number of public and private bodies have developed weather and satellite-based indices to anticipate yield losses. In addition to these simple indices, a multitude of aggregate indices based on statistical methods, machine learning techniques and process-based models have been produced, and little is known about their effectiveness in actually anticipating yield loss. As some of these indices are now frequently used by private companies and public institutions to hedge the risk of crop failure (e.g. through insurance products) it is essential to make a thorough assessment of their accuracy.