金雨菲, 张茂震, 郭含茹, 何卫安. 基于克里格插值与序贯高斯协同模拟的森林碳密度空间估计[J]. 西南林业大学学报, 2013, 33(6): 32-37. DOI: 10.3969/j.issn.2095-1914.2013.06.006
引用本文: 金雨菲, 张茂震, 郭含茹, 何卫安. 基于克里格插值与序贯高斯协同模拟的森林碳密度空间估计[J]. 西南林业大学学报, 2013, 33(6): 32-37. DOI: 10.3969/j.issn.2095-1914.2013.06.006
JIN Yufei1, ZHANG Maozhen1,2, GUO Hanru1, HE Weian3. Comparison of Forest Carbon Spatial Distribution Based on Kriging Interpolation and Sequential Gaussian CoSimulation[J]. Journal of Southwest Forestry University, 2013, 33(6): 32-37. DOI: 10.3969/j.issn.2095-1914.2013.06.006
Citation: JIN Yufei1, ZHANG Maozhen1,2, GUO Hanru1, HE Weian3. Comparison of Forest Carbon Spatial Distribution Based on Kriging Interpolation and Sequential Gaussian CoSimulation[J]. Journal of Southwest Forestry University, 2013, 33(6): 32-37. DOI: 10.3969/j.issn.2095-1914.2013.06.006

基于克里格插值与序贯高斯协同模拟的森林碳密度空间估计

Comparison of Forest Carbon Spatial Distribution Based on Kriging Interpolation and Sequential Gaussian CoSimulation

  • 摘要: 以浙江省仙居县为例,利用普通克里格插值法和序贯高斯协同模拟法对森林碳储量(地上部分)的空间分布进行估计,利用交叉检验方法对其结果进行对比分析。结果表明,仙居县的森林碳密度分布差异较大,大部分地区森林碳密度较低;普通克里格法未能体现森林碳密度的空间差异,具有明显的“平滑”效应,序贯高斯协同模拟在减少平滑影响方面优于前者,序贯高斯协同模拟法的预测结果较克里格插值法高。

     

    Abstract: The spatial distribution of aboveground forest carbon storage in Xianju County, Zhejiang Province was estimated and simulated by means of applying Ordinary Kriging interpolation method and sequential Gaussian cosimulation method,and the outcomes were comparatively analyzed. The results showed that the aboveground forest carbon storage of Xianju County was relatively low and it was unevenly distributed. The Ordinary Kriging interpolation method greatly changed the spatial distribution of aboveground forest carbon storage, which showed a distinctly smooth effect. The sequential Gaussian cosimulation method was better than Ordinary Kriging interpolation method in reducing the smooth affect. For the spatial distribution, Ordinary Kriging interpolation method had to recover the nonforest land prior to representing its distribution,but the sequential Gaussian cosimulation method could represent distribution of the nonforestland without having recovered it. And for the prediction accuracy, the predicted result from sequential Gaussian cosimulation method was relatively higher than that of the Ordinary Kriging interpolation method.Therefore, the sequential Gaussian cosimulation method should be a better and more effective method for studying the forest carbon spatial distribution.

     

/

返回文章
返回