Abstract:

Based on meteorological observation data and rice yield records from Zhalaite Banner from 1969 to 2025, this study employed the 5-year moving average method to separate trend yield from meteorological yield. Using the Mann-Kendall test and multiple regression analysis, it systematically investigated the evolution patterns of thermal, water, and light resources and their quantitative impacts on rice yield, and constructed a yield prediction model. The results show that: (1) the climate of Zhalaite Banner exhibits a significant warming trend, with the annual average temperature increasing by 0.4℃ per decade, the frost-free period extending by 3.6 days per decade, and the ≥10℃ accumulated active temperature increasing by 86.2℃·d per decade. (2) The actual rice yield shows an extremely significant upward trend (increasing by 1380.8 kg/hm² per decade), growing approximately 8.4 times over 57 years, mainly driven by technological progress. However, the meteorological yield fluctuates violently (CV=38.5%), with bumper years, normal years, and yield-reduction years accounting for 29.1%, 27.3%, and 43.6%, respectively. (3) The established quadratic polynomial model can explain 88.44% of the variance in meteorological yield. Validation using independent samples from 2024 to 2025 shows that the relative prediction error is less than 0.5%. The conclusion holds that the increase in thermal resources has significantly contributed to the improvement of trend yield by extending the growth period and expanding the layout of high-yield varieties. However, there is a nonlinear relationship between heat and meteorological yield (positively correlated in the low-to-moderate heat range but negatively correlated beyond the threshold). Coupled with large interannual fluctuations in water conditions and frequent extreme climate events, these factors constitute the main risks to yield stability. Technological progress is the dominant factor in yield enhancement but faces the risk of diminishing marginal returns. It is recommended to optimize variety layout, strengthen water management, and improve early warning systems to enhance the adaptability of rice production to climate change.

Key words: climate change, rice yield, meteorological yield, trend yield, climatic resources