Publication
 

自动曝光对半球摄影法测量叶面积指数及其季节变化的影响 / Influence of automatic exposure on the estimation of seasonal variations in leaf area index measured by digital hemispherical photograph

自动曝光是影响半球摄影法(DHP)测量叶面积指数(LAI)精度的重要误差源之一.本研究基于小兴安岭地区的阔叶红松林、白桦次生林、红松人工林和兴安落叶松人工林,利用DHP和LAI-2200植物冠层分析仪分别测量6—9月每个月中旬的LAI,首先比较两种方法测量LAI的差异性,再检验森林类型和测量时期对建立两种方法测定值间的相关关系是否存在显著影响,最后构建适于校正不同森林类型不同时期自动曝光对DHP测量LAI产生误差的经验模型.结果表明: 4种森林类型4个时期内,在自动曝光设置下DHP测量的LAI比LAI-2200测量值低估20%~49%;森林类型对构建两种方法测量LAI值的经验模型不存在显著影响,而测量时期存在显著影响.本研究构建的A、B两种分类经验模型,分别适用于校正4种森林类型在6和9月、7和8月DHP测量的LAI.经分类经验模型校正后,DHP测量4种森林类型4个时期的LAI值提高了45%~79%,测量精度可提高到83%~94%.通过DHP和LAI-2200测量LAI值间的经验模型,可有效校正自动曝光对DHP测量LAI的影响,极大地提高其测量精度,为使用DHP快捷、高效地测量不同森林类型的LAI及其季节动态提供技术支持.

Automatic exposure is one of the important error sources during measurement of leaf area index (LAI) by digital hemispherical photography (DHP). This study was conducted in a mixed broadleaved-Korean pine (Pinus koraiensis) forest, a secondary birch (Betula platyphylla) forest, a Korean pine plantation and a Dahurian larch (Larix gmelinii) plantation in the Xiaoxing'an Mountains. LAI was measured using DHP and LAI-2200 plant canopy analyzer in the middle of June to September. We compared LAI values measured through these two methods, and then tested whether the forest type and study period had a significant influence on the correlations between the measured values of those two methods. We constructed empirical models for correcting the errors caused by automatic exposure for LAI values measured through DHP at different study periods in different forest types. The results showed that LAI from DHP was underestimated by 20%-49% relative to that from LAI-2200 in four study periods of the four forest types. Forest type had no significant effect on the construction of empirical models between these two measuring methods of LAI, whereas study period showed significant effects. Two classified empirical models (A and B) were constructed, which were suitable for correcting the LAI from DHP in June and September, July and August in four forest types, respectively. After being corrected by the classified empirical models, LAI from DHP of the four forest types increased by 45%-79%, and the measurement accuracy could be improved to 83%-94%. Classified empirical models between LAI from DHP and LAI-2200 could effectively correct the influence of automatic exposure on DHP and greatly improve its measurement accuracy, and provide a technical support for rapid and effective measurement of seasonal changes of LAI in different forest types.

Authors: 
苑振皓, 金光泽, 刘志理 / Z.H. Yuan, G.Z. Jin, & Z.L. Liu
Journal: 
应用生态学报 / Journal of Applied Ecology
Year: 
2018
Volume: 
29
Issue: 
12
Pages: 
4004-4012
DOI: 
10.13287/j.1001-9332.201812.017