研究目的
To analyze the effects of atmosphere and mixed pixels on algorithms for estimating green tide biomass, establish an insensitive algorithm, and discuss factors driving interannual variability.
研究成果
The EVI-based algorithm is effective for estimating Ulva prolifera biomass with low sensitivity to atmospheric and mixed-pixel effects, allowing direct use with top-of-atmosphere reflectance. Application to MODIS data from 2007 to 2016 showed interannual variability, with nutrients from the Sheyang River and Porphyra cultivation in Jiangsu Province being key drivers, while sea surface temperature had no significant impact.
研究不足
The algorithm may not apply during early growth or end stages of green tide due to changes in pigment content, and it cannot estimate biomass for nearshore accumulations where reflectance saturates. Uncertainties exist from statistical and remote sensing data used in factor analysis.
1：Experimental Design and Method Selection:
A water tank experiment was conducted to measure reflectance spectra of Ulva prolifera with different biomass per unit area using a spectrometer. The 6S atmospheric radiative transfer model was used to simulate top-of-atmosphere and Rayleigh-corrected reflectances under various aerosol optical depths. A linear mixing model simulated mixed-pixel reflectance.
2：Sample Selection and Data Sources:
Ulva prolifera samples were collected from coastal waters off Qingdao, China, with biomass per unit area ranging from
3：44 to 06 kg/m2. MODIS L1B daily data from 2007 to 2016 were used for long-term analysis. List of Experimental Equipment and Materials:
A water tank covered with black plastic membrane, FieldSpec4 ASD spectrometer, electronic balance, filter papers, seawater, and Ulva prolifera samples.
4：Experimental Procedures and Operational Workflow:
Samples were weighed, placed in the tank, and reflectance was measured. Reflectance data were integrated to MODIS bands using spectral response functions. Simulations for atmosphere and mixed pixels were performed.
5：Data Analysis Methods:
Exponential relationships between biomass and spectral indices (e.g., EVI, NDVI) were analyzed using R2 and average percentage deviation (APD). Correlations with environmental factors were assessed.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
