- 期刊
- OpenAccess
Decisions Affecting Estimations of Understory Light Environments during Photograph Acquisition, Storage, and Analysis Using Hemispherical Photography
半球面影像攝取、貯存及分析過程中一些決定對林下光環境估測之影響
摘要
利用半球面影像技術來探討林冠結構與林下光環境已有約半世紀。一些研究檢驗了此一技術從影像攝取、儲存、到分析過程的一些潛在問題及限制。一般使用者必需清楚這些問題對其研究的影響，但這些問題及限制之克服有些要依賴技術的改良精進。本研究探討相機類型、影像解析度、影像壓縮及不同分析軟體對估測林冠光穿透的影響。研究結果顯示影像壓縮與否對分析之結果並無顯著影響而解析度則有顯著影響，解析度越高所估算之林下光照越高。未壓縮影像之貯存要花相當久的時間，而適合半球面影像拍攝之均質無直射光天空狀態卻相當有限。建議研究者使用高解析度壓縮之格式來貯存影像，因為在林冠較為鬱閉的森林，高解析度較能區分細小的林冠成分與孔隙。使用傳統底片相機及數位相機所推估之林冠光穿透可相差超過5%，不同影像分析計算軟體所造成之差異亦可達5%以上，因此跨研究之比較必需注意所使用之解析度、相機系統以及分析軟體是否相同。如果兩研究間林冠光穿透之差異小於10%，則在解釋結果時應特別小心。本研究亦發現散射光穿透率之估測受上述各項因子之影響較直射光穿透率大。這是由於半球面影像之中心通常未受遮蔽，故不易受上述因子之影響，而在亞熱帶地區此部分之影像大都位在太陽軌跡上，故對直射光穿透之影響較大。
並列摘要
Hemispherical photography has been used to study forest canopy geometry and understory light environments for approximately half a century. Several studies have examined some potential problems and limitations of hemispherical photography from image acquisition and storage to analysis. General users of hemispherical photography should be aware of situations when these problems might have major impact on their studies, and of solutions to many of the problems, which rely on advances in techniques and associated computer technology. In this study, we examined the influences of camera type, resolution, compression, and the image analysis package on the estimation of canopy light transmittance using hemispherical photography. The results indicate that compression did not affect the estimation of canopy light transmittance in our studied forest but resolution did; higher resolutions yielded higher estimates of canopy light transmittance. We found that storing uncompressed images is much more time consuming, which can be a disadvantage in the field when the time-period suitable for hemispherical photograph acquisition is limited. We recommend the use of compressed high-resolution images in forests with dense canopies because they provide better distinction between small plant components and gaps. Digital and film camera systems differed in estimates of canopy light transmittance by more than 5%, as did different analytical packages. Any cross-study comparison must look into the resolution, software, and camera system being used. If canopy light transmittance differs by less than 10%, researchers should interpret their results with great caution. We also found that estimates of indirect light transmittance were less robust than direct light transmittance. This is because there is less uncertainty in the calculation of transmittance for the center of a hemispherical photograph which at subtropical latitudes is mostly located in the sun path and contributes more to direct than indirect light transmittance.