模擬衛星遙測估算甘蔗葉片葉綠素含量

Mimicking Satellite Remote Sensing of Chlorophyll Content in Sugarcane (Saccharum officinarum) Leaves

10.30061/CEB.200506.0006

黃文達(Wen-Dar Huang)；許明晃(Ming-Huang Hsu)；楊志維(Zhi-Wei Yang)；陳建璋(Jane-Chang Chen)；蔡養正(Yang-Zenq Tsai)；張新軒(Shin-Shinge Chang)；楊棋明(Chi-Ming Yang)

甘蔗 ； 反射率 ； 衛星遙測 ； 植生指數 ； 葉綠素 ； 花青素 ； Sugarcane ； Satellite remote sensing ； Vegetation indices ； Chlorophyll ； Anthocyanin

作物、環境與生物資訊

2卷2期（2005 / 06 / 01）

137 - 147

繁體中文

本研究以不同生育期第60、90、360及540天之甘蔗(Saccharum officinarum L.)葉片為材料，進行葉片反射光譜測定，並以反射率模擬計算衛星遙測植生指數(vegetation index, VI)，估算其葉片之葉綠素(chlorophyll, Chl)含量。隨著生育日數的增加，甘蔗葉片葉綠素及類胡蘿蔔素含量逐漸減少。由反射光譜分析顯示，甘蔗葉片在550nm與710nm之反射率對色素含量變化具有最大敏感度。以近紅外光波段750nm反射率分別與710nm及550nm反射率計算植生指數SRVI(simple ratio vegetation index)、NDVI(normalized difference vegetation index)與GI(greenness index)、NDVI(下標 Green)，以及模擬SPOT衛星寬波段平均反射率計算植生指數SRVI(下標 broad)及NDVI(下標 broad)後，與葉片色素含量進行迴歸分析，可得到葉綠素含量估算方程式。以此六組方程式估算正常生長之葉片葉綠素含量，其準確度相當高；但估算受逆境含花青素之葉片葉綠素含量時，則有明顯之誤差。本研究結果顯示利用反射光譜計算植生指數，非破壞性地估算正常生長狀況下之葉片葉綠素含量之變化，進而監測作物生長狀況，為有效可行的方法。

The reflectance spectra of sugarcane (Saccharum officinarum) leaves at 60, 90, 360 and 540 days after planting were measured to calculate vegetation indices to mimic the satellite remotely estimate of chlorophyll content. The contents of pigments declined following the increasing of days after planting. The signature analysis of reflectance spectra indicated that the maximum sensitivity to chlorophyll content in the leaves of sugarcane was at 550nm and 710nm. SRVI (simple ratio vegetation index) and NDVI (normalized difference vegetation index) calculated using reflectance at 750nm and 710nm and GI (greenness index) and NDVI(subscript Green) calculated using reflectance at 750nm and 550nm correlated very well with chlorophyll content. SRVI(subscript broad) and NDVI(subscript broad) calculated using simulated broad-band reflectance of SPOT satellite also correlated well with chlorophyll content. The developed algorithms predicting leaf chlorophyll from the leaf optics were validated from two independent data. A significant error in chlorophyll prediction of the stressed leaves containing anthocyanin was observed. Results suggest that it is feasible to create indices using reflectance spectra for non-destructive estimation of chlorophyll content, by which to indirectly monitor crop growth under normal environment.