Data from national meteorological stations in 90 counties of Jiangxi province during 2012-2021 and 8 microclimate stations in rice fields during 2017-2021 were used to analyze the differences in temperature between rice fields and stations, and the statistical results based on H1 (NS)(daily max temperature of national stations ≥35 °C), H1 (RS)(daily max temperature of rice fields ≥35 °C), H2 (NS)(daily max temperature of national stations ≥37°C) and H2 (RS) (daily max temperature of rice fields ≥ 37 °C ) were compared. Furthermore, proposing high-temperature damage indicators more suitable for early rice in Jiangxi, which would promoting capabilities of high-temperature monitoring and defense. The results showed that: (1) compared with the national stations, the rice fields had a cooling effect, and the Tmax(daily maximum temperature), T(daily average temperature), and Tmin(daily minimum temperature) were 0.6 °C, 0.4 °C, and 0.3 °C lower than national stations. What's more, the cooling effect increased with the growth of early rice, which showed as heading-maturity period>transplanting-booting period>sowing-transplanting period. (2) The cooling effect was higher in sunny days than in cloudy and rainy days during the high temperature period, and T, Tmax and Tmin of early rice field were 0.7 °C, 0.8 °C and 0.6 °C lower than that of the national stations. On cloudy and rainy days, the decrease in temperature of rice fields was similar. (3) There were significant differences in daily and hourly air temperature between 6 rice fields and the corresponding national stations during H1 (NS) heat damage occurred, and the cooling intensity of the rice field was greater at night than in the daytime. The Tmax, T and Tmin of the 6 fields were generally about 1.2 °C lower, which was strongest in Jinxi filed with T, Tmax and Tmin was 1.5 °C, 1.1 °C and 2.6 °C lower. (4) H1 (NS) was the largest and H2 (RS) was the smallest in terms of the number of counties affected by heat damage, the average duration, the average yield reduction simulation value of affected counties and the simulation value of the whole in recent 10 years based on different heat damage index. There was no significant difference in the result of average duration, the simulated average yield reduction rate of affected counties, and the simulated yield reduction rate of the whole based on H2 (NS) and H1 (RS), while they were all different with the result based on H1 (NS) or H2 (RS). And the simulated value of the reduction rate of the whole calculated by H2 (NS) was closer to the actual value. Therefore, national stations Tmax ≥37°C was more suitable as an index of heat damage for early rice in Jiangxi. [ABSTRACT FROM AUTHOR]