摘要:全球暖化已經從先前人云亦云的揣測,到現在可以親身感受到無須爭辯的事實,暖化對於生態系統有相當大的衝擊,尤其是對氣候變化極為敏感且脆弱的熱帶與亞熱帶的濕潤森林。造成暖化最主要的成因為人類活動行為將大量二氧化碳排放至大氣中,提高其溫室氣體的濃度。然而地球的自我調適可以藉由植物的光合作用將二氧化碳吸存回陸地,此機制可以減輕溫室效應對於生物圈的衝擊以及減緩環境惡化的趨勢。評估森林吸存能力最重要的指標為初級生產淨量(簡稱NPP),然而量化這指標無論是地面調查或電腦模式模擬,在技術上都有非常大的挑戰,而對於研究地形崎嶇的臺灣山地森林更有著很大的限制。異素生長定律為經典生態學理論,依據這定律,森林NPP可以藉由單一森林結構功能特徵(如:生物量、葉面積指數)或生理特性(例:葉片養分、色素、含水量、光合作用能力)來估測,而許多這些森林功能與生理特性可以從衛星遙測的光學波段所運算出的強化型植生指數(Enhanced Vegetation Index, EVI)來間接推估。因此,本一年期研究計畫將結合地面所量測的光學遙測光譜、異素生長定律與地面所蒐集的NPP資料來評估運用美國最新一代大地衛星作業地表成像儀(Landsat-8 Operational Land Imager,簡稱Landsat-8)所運算出的EVI估算臺灣東北部濕潤森林NPP的可行性。本研究計畫包含2個0.6公頃的闊葉(福山試驗林)與針葉(棲蘭山長期生態研究區)林。在地面資料蒐集方面,計畫主持人將每月到研究樣區蒐集地面NPP資料,包含樹木碳累積量、樹冠生產量以及地下部細根生產量。在光學遙測分析方面,為了減輕樣區多雲霧的研究限制,計畫主持人將每月至樣區採集少量優勢種喬木葉片,並使用可攜帶式高光譜輻射儀配合光學積分球體儀量測樣區具代表性葉片反射率,再運用冠層輻射轉換模式模擬森林樹冠高光譜反射率,之後主持人將使用光譜折合計算把高光譜資料轉成Landsat-8光譜波段並將反射率轉換成EVI,這資料可用來導入異素生長定律模式來計算每月森林NPP與EVI之間的相對關係,且利用地面同時期所蒐集之NPP資料來做模式校正將這相對關係轉換成絕對數值,用來準確估測出臺灣東北部濕潤森林每月的NPP。受惠於美國大地衛星傳承計畫持續長期全球連續性的地表觀測,本研究成果將可促進系統化長期全球尺度高空間解析度濕潤森林新陳代謝監測的應用。
Abstract: The recognition of global warming has been transformed from an anecdote to an unambiguous fact. The ramifications of elevated temperature on ecosystems are pronounced, especially for tropical and subtropical humid forests, which are extremely sensitive and vulnerable to climate fluctuation. The main driver resulting in global warming is the acceleration of anthropogenic activity induced carbon emission to the atmosphere, which significant raises the concentration of greenhouse gases. Forests can sequester and store a substantial amount of carbon back to the land via photosynthesis, which is an earth adaption mechanism to mitigate the impacts of global warming on the biosphere and to retard the rate of environmental degradation. However, the quantative assessments of forest carbon uptake (also known as net primary production [NPP]) at different scales from field inventory to computer modeling are technically challenging, especially for the Taiwan’s montane forests where are situated on the rugged terrain. The allometric scaling law (ASL) is a classical ecology concept. According to ASL, forest NPP can be derived using one single functional (e.g., biomass, leaf area index) or physiological (such as the leaf nutrients, pigments, leaf water content or photosynthesis rate) trait, and many of these attributes may be indirectly measurable using the Enhanced Vegetation Index (EVI), which is commonly derived using optical spectral bands acquired from a satellite sensor. Therefore, the main objective of this one-year project is to integrate field spectroscopy, ASL and field NPP observation to assess the feasibility of utilizing the EVI derived from the Landsat-8 Operational Land Imager (hereafter “Landsat-8”) to estimate NPP of humid forests in the northeastern Taiwan. This proposed study will include two 0.6 ha broadleaf (Fushan Experimental Forest) and conifer (Chilan long-term ecological research site) forests. Field NPP of the forests, consisting of woody carbon accumulation, and canopy and fine root productions, will be measured monthly. Since the study region is very humid with a high frequency of cloud and fog coverages, it is impractical to derive monthly Landsat-8 EVI from the sites. Therefore, a portable spectroradiometer with an integrating sphere will be utilized to measure the leaf spectra of the dominant tree species, which will be converted to canopy reflectance by canopy radiative transfer. These hyperspectral data will be convoluted to Landsat-8 bands, which will be used to calculate the EVI. The vegetation index will be injected into ASL to derive the relative relationship between the modeled monthly EVI and NPP. Field collected monthly NPP data will then be applied to calibrate the model and to transform the relative relationship to actual measures. This will enhance the model performance for precise monthly NPP estimation. The Landsat program (as known as the Landsat Legacy Program) provides wall-to-wall global terrestrial observation for several decades. Therefore, the outcomes of the proposed project will facilitate the applications of long-term systematic monitoring of the metabolism of global humid forests at the fine resolution.