Estimating the impact of ground ozone concentrations on crop yields across China from 2014 to 2018: A multi-model comparison

Highlights: • Ozone induced yield losses in China from 2014 to 2018 were estimated using a diversity of O3-crop models. • The total Relative Yield Losses (RYL) of wheat and rice in China was 18.4%–49.3% and 6.2%–52.9%. • Large variations existed between different regional O₃-crop models when applying to large scales. • It is a preferable strategy to employ suitable corresponding O₃-crop models in different regions. • A systematic framework with massive regional O₃-crop models is required. Ground level ozone exerts a strong impact on crop yields, yet how to properly quantify ozone induced yield losses in China remains challenging. To this end, we employed a series of O₃-crop models to estimate ozone induced yield losses in China from 2014 to 2018. The outputs from all models suggested that the total Relative Yield Losses (RYL) of wheat in China from 2014 to 2018 was 18.4%–49.3% and the total RYL of rice was 6.2%–52.9%. Consequently, the total Crop Production Losses (CPL) of wheat and rice could reach 63.9–130.4 and 28.3–35.4 million tons, and the corresponding Total Economic Losses (TEL) could reach 20.5–44.7 and 11.0–15.3 billion dollars, stressing the great importance and urgency of national ozone management. Meanwhile, the estimation outputs highlighted the large variations between different regional O₃-crop models when applying to large scales. Instead of applying one unified O₃-crop models to all regions across China, we also explored the strategy of employing specific O₃-crop models in corresponding (and neighboring) regions to estimate ozone induced yield loss in China. The comparison of two strategies suggested that the mean value from multiple models may still present an inconsistent over/underestimation trend for different crops. Therefore, it is a preferable strategy to employ corresponding O₃-crop models in different regions for estimating the national crop losses caused by ozone pollution. However, the severe lack of regional O₃-crop models in most regions across China makes a robust estimation of national yield losses highly challenging. Given the large variations between O₃-crop interactions across regions, a systematic framework with massive regional O₃-crop models should be properly designed and implemented.