基于改进的像元二分模型估测郁闭度

目的郁闭度是森林资源调查的一个重要因子，它不仅可以反映森林冠层的郁闭程度和树木利用空间的程度，并且能够指示林分密度。遥感为区域和全球尺度精确估测郁闭度提供了前所未有的契机，使得大面积郁闭度监测制图成为可能。本文旨在利用资源三号卫星影像数据结合模型对区域尺度上落叶松林进行郁闭度估测。方法以黑龙江省佳木斯市桦南县孟家岗林场落叶松人工林为研究对象，首先采用像元二分模型对植被覆盖度进行估算，通过探寻植被覆盖度与郁闭度的关系，对像元二分模型进行改进，通过计算不同累积频率归一化植被指数（NDVI）的值作为模型参数，对比不同参数取值时模型拟合效果得到最优模型，并利用模型估测郁闭度。结果拟合结果表明，2%累积频率下模型拟合效果最好，模型R2为0.838，RMSE为0.045，最后利用该模型得出孟家岗林场落叶松人工林郁闭度分布图。结论利用改进的像元二分模型可以较为准确地估测郁闭度。本研究对探索我国北方落叶松人工林郁闭度遥感估测提供了更有效的途径，同时也为森林资源和参数的调查提供一定的参考依据。 

Estimation of canopy closure based on improved dimidiate pixel model

ObjectiveThe canopy closure is an important factor in forest resource investigation. It can not only reflect the closure degree of forest canopy, the degree of forest using space, but also indicate the density of the forest. Remote sensing provides an unprecedented opportunity for accurate estimation of canopy closure at regional and global scales, making it possible to monitor and map large area canopy closure. This study aims to estimate canopy closure of Larix gmelini on regional scale through the ZY-3.MethodIn this study, Larix gmelini in Mengjiagang of Huanan County, Jiamusi City of Heilongjiang Province, northeastern China was taken as the research object. Firstly, dimidiate pixel model was used to estimate vegetation coverage. By exploring the relationship between vegetation coverage and canopy closure, the dimidiate pixel model was improved. The normalized difference vegetation index (NDVI) was calculated by calculating the cumulative frequency as model parameters. The optimal model was obtained by comparing the model fitting effect of different parameters, and the canopy closure was estimated by the optimal model.ResultThe fitting results showed that the fitting effect of 2% cumulative frequency was the best, with R2 was 0.838 and RMSE was 0.045. Finally, the model was used to obtain the distribution map of the canopy closure of Larix gmelini in Mengjiagang.ConclusionThe results show that improved dimidiate pixel model can get accurate estimation of canopy closure. This study provides a more effective way to estimate the canopy closure of Larix gmelini in the north of China, and also provides a reference for the investigation of forest resources and parameters.