Abstract: | 應用MODIS觀測資料推估熱帶雨林淨初生產力 Estimating Net Primary Productivity (NPP) on Tropical Rain Forest using MODIS Observation Data 欲評估森林生態系中之碳循環模式, 淨初生產力 (Net Primary Productivity, NPP) 被稱為一項重要的指標。然而,由於熱帶雨林面積廣大及缺乏地表實測資料的因素,要了解森林生態系如何藉由地表物質來進行循環,是一項困難的任務。因此,我們應用遙感探測資料推估熱帶雨林NPP之三個重要的參數,包括光利用效率 (Light Use Efficiency, LUE)、 光合作用有效輻射比率 (Fraction of Photosynthetically Active Radiation, FPAR) 和光合作用有效輻射 (Photosynthetically Active Radiation, PAR)。 . 基本上, Montieth 方程式是用來測量熱帶雨林地上之 NPP 量。Rahman 等人(2004)提出微氣象模式, 使用 MODIS 海洋波段所演算出的光化學反射率指標 (MODIS Photochemical Reflectance Index, MODPRI), 針對陸域植生進行”連續野外”的LUE推估。 FPAR 是由常態化植生指標 (Normalized Difference Vegetation Index, NDVI) 演算而得。PAR 是從45%的入射短波太陽輻射 (Rs) 取得。 在這篇文研究中, 我們應用 2000~2006 年的 MODIS 衛星影像資料所推出之結果, 和 NASA 的 MOD17 產品進行比較。在參考資料方面, 本文所採用的 NPP-APAR 資料中,APAR 是來自 Bukit Soeharto (BKS) 微氣象站的地真資料 (2001和2002年)。 . 本研究主要目的是, 藉由 MODIS 遙測影像資料, 在沒有使用微氣候資訊的前提下, 發展一種推估 NPP 的新方法。 . 結果顯示出, 以 NPP Rs、 NPP MOD17 和 NPP APAR 的方法推估婆羅洲 NPP 量,在 2001 年分別為 1076.546 gC m2/y、3028.850 gC m2/y 和 1615.509 gC m2/y, 2002 年分別為 1522.324 gC m2/y、2890.467 gC m2/y 和 1987.113 gC m2/y。 其中, NPP Rs 和 NPP MOD17 之間沒有顯示出相關性; 相同的情形也發生在 NPP APAR NPP Rs 和 NPP APAR 的結果顯示出,在 2001 年和 2002 年的資料趨勢上有良好的相關性。 . 從這個結果來看,我們同時可以了解到:在這研究區域中,NASA 提供的 NPP 資料有高估的情況,而 NPP Rs 和 NPP APAR 相比較時, 雖然 NPP 以平均來看仍然是低估的情況, 但結果比 NASA 所提供的結果好。這結果也提供我們應用遙測影像資料推估 NPP 的一項新方法。 隨著先前研究運用本研究結果, 將可以在推估 NPP 時上得到較準確的結果。Estimating Net Primary Productivity (NPP) on Tropical Rain Forest Using MODIS Observation Data Net Primary Productivity (NPP) is known as an important index for evaluating the carbon cycle in forest ecosystem. However, due to the factors of large area and lack of ground measurement data in tropical forest, it is hard to understand how forest ecosystem runs by land materials. Therefore, we use remote sensing data to estimate the three important parameters of NPP in tropical rain forest such as Light Use Efficiency (LUE), Fraction of Photosynthetically Active Radiation absorbed by vegetation (FPAR), and Photosynthetically Active Radiation absorbed by vegetation (PAR). Basically, the Monteith equation is used to measure aboveground NPP in the tropical forest. Rahman et al. (2004) proposed a micrometeorological model to estimate ‘‘continuous field’’ LUE for terrestrial vegetation using MODIS Photochemical Reflectance Index (MODPRI) from the MODIS ocean bands. FPAR is calculated from Normalized Difference Vegetation Index (NDVI). PAR is from 45 % of incoming shortwave solar radiation (Rs). In this study, we use MODIS data from 2000 to 2006 and compare the results with MOD17 of NASA Product. Absorbed Photosynthetically Active Radiation (APAR) from in-situ data (2001 and 2002) of Bukit Soeharto (BKS) site of micrometeorological station is used to produce a NPP APAR. Then, it’s a reference data in this study. The main objective of this study is to get a new method to estimate NPP from MODIS remotely sensed data without using micrometeorological information over the study areas. This method creates a challenge to monitor ecosystem productivity in global area and even in difficultly-accessed regions. The results show estimated value for the distribution of NPP over Borneo Island by three methods : NPP Rs, NPP MOD17 and NPP APAR. These values are 1076.546 gC m-2 y-1; 3028.850 gC m-2 y-1; 1615.509 gC m-2 y-1 in 2001, and 1522.324 gC m-2 y-1; 2890.467 gC m-2 y-1; 1987.113 gC m-2 y-1 in 2002, respectively. NPP Rs and NPP MOD17 values showed that there is no correlation between them. NPP Rs and NPP APAR outcomes represent good correlation in the trend of 2001 and 2002 data. From this result we can conclude that: the NPP data provided by NASA is overestimated in this region, NPP Rs gives better result in comparison with NPP APAR, although it is still underestimated. Improved methods for estimating NPP by using MODIS observation data is needed to be developed. With advanced study employing the methodology of this study, better accuracy in estimating NPP can be achieved. |