In Situ Hyperspectral Data Analysis for Canopy Chlorophyll Content Estimation of an Invasive Species Spartina Alterniflora Based on Prosail Canopy Radiative Transfer Model

Jinquan Ai,Wei Gao,Runhe Shi,Chao Zhang,Zhibin Sun,Wenhui Chen,Chaoshun Liu,Yuyan Zeng
DOI: https://doi.org/10.1117/12.2186973
2015-01-01
Abstract:Spartina alterniflora is one of the most serious invasive species in the coastal saltmarshes of China. An accurate quantitative estimation of its canopy leaf chlorophyll content is of great importance for monitoring plant physiological state and vegetation productivity. Hyperspectral reflectance data representing a range of canopy chlorophyll content were simulated by using the PROSAIL radiative transfer model at a 1nm sampling interval, which was based on prior knowledge of S.alterniflora. A set of indices was tested for estimating canopy chlorophyll content. Subsequently, validation were performed for testing the performance of indices, based on the PROSAIL model using in situ data measured by a Spectroradiometer with spectral range of 350-2500nm in a late autumn in a sub-tropical estuarine marsh. PROSAIL simulations showed that the most readily available indices were not good to be directly used in canopy chlorophyll estimation of S.alterniflora. The modified Chlorophyll Absorption in Reflectance Index MCARI[705,750] was linear related to the canopy chlorophyll content (R-2=0.94), but did not achieve a satisfactory estimation results with a high RMSE (RMSE=0.95 g.m(-2)). We optimized the index MCARI[705,750] by introducing a scale conversion coefficient to the formula to solve data units inconsistent, which is between the practical application unit and the unit used in the process of establishing the index, and balance scale transformation through radiative transfer models and examing corresponding canopy reflectance index values. We proposed index Optimized modified Chlorophyll Absorption in Reflectance Index OMCARI[705, 750]. The results showed that the index OMCARI[705, 750] had higher precision of prediction of chlorophyll for S.alterniflora (R-2=0.94, RMSE=0.41 g.m(-2)).
What problem does this paper attempt to address?