The purpose of this study was to improve utilization rate of climate resources and reasonably optimize layout of winter facility agriculture in Henan Province. Based on daily minimum temperature data from 113 national meteorological stations during 1981-2018 in Henan Province, the mathematical statistics method was used to calculate standard deviation of minimum temperature in winter and climatic probability, frequency and occurrence intensity of different low temperature disaster indices. The low temperature risk index of facility agriculture was constructed by using the equal weight method. The low temperature risk levels of facility agriculture were classified using the optimal clustering method of ordered samples combined with low temperature disaster data, and the risk assessment was conducted using the risk level combined with land-cover data. The results show that the climatic probability and occurrence frequency of low temperature disaster (less than or equal to -5 ℃ ) in all parts of Henan Province are more than 0.80, so it is not suitable for development of small plastic arch sheds, and it is not suitable to use single-layer plastic greenhouses for facility agricultural production. The climatic probability and occurrence frequency of temperature less than or equal to -10 ℃ in both Xinyang and southern Nanyang are less than 0.20, it is suitable for development of plastic greenhouses. The climatic probability and occurrence frequency of temperature less than or equal to -10 ℃ in Hebi, Anyang and Puyang are all more than 0.80, so in these regions solar greenhouses should be developed to ensure crop growth in facilities. In the plastic greenhouse development area, Gushi County and Shangcheng County of Xinyang are mild risk areas for low temperature disasters, most of Xinyang, parts of Nanyang and Zhoukou are moderate risk areas, and all other areas are severe risk areas. In the solar greenhouse development areas, Puyang, Anyang and parts of Hebi are areas with high risk of low temperature disasters. Therefore, the low temperature disaster risk of different levels in Henan Province in winter has a certain regional pattern. Arrangements on planning of facility agriculture should be done in different regions according to local low temperature disaster risks.

The study on the spatio-temporal variation of reference crop evapotranspiration and its causes in Huang-Huai-Hai region is an effective way for improving the utilization efficiency of regional agricultural water resources. Based on the daily observation data at 364 meteorological stations in Huang-Huai-Hai region during 1961-2018, the reference crop evapotranspiration (ET0) was calculated by using FAO Penman-Monteith equation. And on this basis the spatio-temporal variation of ET0 and contribution of influencing factors were analyzed by using trend analysis, sensitivity analysis and contribution rate analysis. The results are as follows: (1) The annual mean ET0 was 978.7 mm in Huang-Huai-Hai region during 1961-2018, and it decreased gradually from the central to surrounding areas. The annual mean ET0 decreased significantly with the rate of 12.5 mm·(10 a)-1 from 1961 to 2018, and the decadal variation of ET0 was obvious, especially before the 1990s. Meanwhile, the annual mean temperature increased significantly, and the wind speed, sunshine hours and relative humidity decreased significantly, but their climatic tendency rates in each province and area of Huang-Huai-Hai region were different slightly. (2) The sensitivity coefficients of ET0 to mean temperature, wind speed and sunshine hours were positive, while that of ET0 to relative humidity was negative. (3) The contribution rates of mean temperature and relative humidity to ET0 were negative, while that of wind speed and sunshine hours were positive, and the absolute value of contribution rate decreased in turn from wind speed, sunshine hours, mean temperature and relative humidity. (4) The decrease of wind speed and sunshine hours was a main reason for ET0 decrease in Huang-Huai-Hai region, but the reason of ET0 decrease was slightly different in each province and area.