基于Landsat 8影像的地温对森林地上生物量的影响

Influence mechanism of land surface temperature to the forest aboveground biomass based on the Landsat 8 image

【目的】森林地上生物量对评价全球碳平衡和气候变化有重要作用,以安徽省霍山县为研究区,Landsat OLI和TIRS影像为基础数据,研究地表温度对森林地上生物量的影响效应。【方法】以Landsat OLI为遥感数据源,在基于灰度共生矩阵的支持下,提取覆盖研究区的遥感影像的纹理特征,并结合原始多光谱遥感影像和研究区的地形特征构建基于3种因子的森林蓄积量反演模型,并根据最优模型来推算研究区的森林地上生物量。根据Landsat TIRS数据对研究区的地表温度进行反演,并将森林地上生物量和地表温度反演数据结合,分析地表温度对森林地上生物量的影响机制。【结果】1)本研究共构建3种模型对森林蓄积量进行反演研究,包括:基于光谱特征的森林蓄积量反演模型,基于纹理特征的森林蓄积量反演模型和基于多特征的森林蓄积量反演模型。2)对森林蓄积量反演模型的精度进行评价,结果显示基于多特征的森林蓄积量反演模型精度最高,复相关系数达到0.7519,说明该模型可用于研究区的森林蓄积量反演研究,森林蓄积量反演结果可用于研究区的森林地上生物量推算。【结论】研究结果表明在研究区内海拔200 m以下和700 m以上的区域,地表温度对森林地上生物量的影响效应不明显;在海拔200~400 m梯度范围内,森林地上生物量与地表温度的相关性逐渐增大;海拔400 m以上,森林地上生物量与地表温度的相关性与200~400 m范围内相比逐渐减小,但总体上,在400~700 m海拔梯度内,两者仍呈现出较强相关性。

【Objective】The forest above ground biomass(AGB)has a significant influence on the evaluation of the global carbon-oxygen balance and climate change,Huoshan county of Anhui province is selected as the study region,the Landsat OLI and TIRS data were used to study on the influence of the land surface temperature(LST)on the AGB.【Method】The Landsat OLI images were processed by the Gray level Co-occurrence Matrix to extract the texture features of the remote sensing dataset.In this study three different inversion model were established to estimates the growing stock volume of the study area,after compared the inverse results,the best estimation model was used to estimate the above ground biomass(AGB)of Huoshan County.The Landsat TIRS was used to inverse the land surface temperature(LST)of the study area,then the relationship between the LST and the AGB of the Huoshan county was analyzed.【Result】1)Three forestry growing stock volume inversion models were established in this study,they are inversion model based on multispectral feature,inversion model based on texture feature,and inversion model based on multiple features,respectively.2)Based on the evaluation results of the inversion models,the inversion model based on multiple features obtained the best accuracy,the R2 of this model is 0.7519.It indicates that this model can be used to inverse the growing stock volume of the study area,the inversion result can be used to calculate the AGB of the Huoshan county.【Conclusion】In the study area,under altitude level of 200 m and above the altitude level of 700 m,the LST has insignificant impact on the AGB.Within the altitude level of 200-400 m,the correlation between AGB and LST is gradually increased.When the altitude is above 400 m,the correlation between AGB and LST is slightly decreased compared with the altitude level of 200-400 m,but under the altitude level of 400-700 m,the AGB and LST is still relatively highly correlated.