Studies on regional difference for thermal influence about forest

本研究的目的是利用卫星数据分析森林砍伐热影响的区域差异。分析如下：1)利用Landsat TM数据计算土地利用与地表温度之间的关系；2)计算1x1 km栅格内土地利用率与地表平均温度之间的多元回归方程。方程表示地表温度的估计，当土地利用发生变化时，计算标准多元回归方程，比较森林砍伐的区域差异。利用Landsat TM数据对北海道（Tomaikomai、Sarobetu）、关头（fuusa）、四国（Kouchi）和九州（Aso）进行分析。结果表明：1)森林地区的地表温度预测值最低，城市地区最高；2)除北海道（Sarobetu）外，所有地区的标准回归系数都是恒定的。我们认识到，森林砍伐的热影响可以使用标准回归系数进行比较。标准回归系数的取值范围为Tomakomai、Kouchi和Fukusa。

The purpose of this study is the regional difference for thermal influence of deforestation usig satellite data. The analyses are performed as follows ; 1) the relationship between land-use and surface temperature are calculated using Landsat TM data, 2) multi-regression equations between the rate of land-use for 1x1 km mesh and mean surface temperature for 1x1 km mesh are calculated. the equations express estimation of surface temparture, when land-use change, 3) standard multi-regression equations are calculated for comparison between the regional difference of deforestation. Landsat TM data are used for the analysis about Hokkaido (Tomaikomai, Sarobetu), Kantou (Fukusa), Sikoku (Kouchi) and Kyuushuu (Aso).The results are obtained that 1) the estimated surface temperature is the lowest on forest and the highest on city areas, 2) standard regression coefficients are constant for Kantou (Fukusa) and Sikoku (Kouchi) through all data but Hokkaido (Sarobetu). We recognized that thermal influence of deforestation is able to compare usig standard regression coefficients. Tlie range of standard regression coeficient is Tomakomai, Kouchi and Fukusa.