Abstract: In this article, the authors at first made three-principled selection, partial correlation, best multi-prin-ciple regression selectlon in sequence,Secondly on the basis of residual analysis, the authors made test ofhigh leverage case, influence case and outlier on sample data and then got rid of the unusual samples byregression diagnostics.Finally the authors establishe regres；ional equation between belt porosity and beltprotective characteristics, After the model was qualialified by test in outside practice,the authors took fur-ther probe into the applicatlons such as the design of belthilding, broadcabting of porosity, belt qualitytest, best regulation of belt structure and best gradual multi-purpose comblnation of belt protective characteristics.

Key words: canopy temperature, infrared thermometer, Quercus variabilis, micrometeorological factor.