Monitoring and Predicting of Maize Chilling Damage Based on Crop Growth Model in Northeast China

Abstract

Although atmospheric temperature is obviously increased in northeast China with the global warming, regional maize chilling damage still occurs due to North extension of cultivation region of relative late-maturing varieties. Therefore, the research on monitoring and predicting maize chilling damage is still necessary. In this paper, we firstly constructed chilling damage indexes based on northeast China maize growth model (NEC_MaGM), and then developed the methods of monitoring and predicting maize chilling damage. The results are as follows: (1) In the eight individual chilling damage indicators selected from adverse weather conditions and the response of maize growth to low temperature, the second one (DC_Tas9, reduction of heat units during tasseling to September 31 compared with multi-year average) and the first one (DN_Tas, the difference of tasseling stage between this year and multi-year average) had the best historical matching capability for chilling damage. The accuracy of chilling damage simulation based on DN_Tas and DC_Tas9 was 93.0% and 81.4%, respectively. (2) Based on formation mechanism of chilling damage, historical matching capability of every individual indicator and its independence, we constructed an integrated chilling damage index, which included the first, the second, the forth (DW_GrS, the loss of WSO suffered by low temperature in growing season compared with multi-year average) and the seventh individual indicators (DW_Fro, the loss of WSO suffered by first frost). (3) Chilling damage was monitored for every site by using NEC_MaGM and indicators, and then regional maize chilling was described by calculating the proportion of the site with chilling damage. On the base of integrated index, we defined that regional chilling damage should have more than 45% of maize chilling site. Thus, the accuracy and the threat score of maize chilling damage simulation were 93.6% and 84.2%, respectively. The result of independent samples test was consistent with actual situation. (4) Monitoring of chilling damage in the grid scale could be described more detailed spatial distribution. With the constantly updated live weather data, chilling dynamic monitoring could be achieved. The causes of chilling damage in the different regions were explored by using every individual indicator. The severity and spatial dis-tribution of chilling damage simulations in the grid scale were fairly consistent with the literature. This method of monitoring chilling damage was favorable to business development of agricultural meteorological services. (5) According to weather data measured and predicted by regional climate models, combined with climate average data, regional maize chilling damage could be predicted by using NEC_MaGM. Case study showed that the method reflected to some extent the development and severity of chilling damage, but its accuracy was not only related with crop models, but also depended on simulation ability of regional climate model. The study could provide a scientific basis for the disaster prevention and mitigation.