Publications

Zheng, Y; Zhang, M; Wu, BF (2016). Using high spatial and temporal resolution data blended from SPOT-5 and MODIS to map biomass of summer maize. 2016 FIFTH INTERNATIONAL CONFERENCE ON AGRO-GEOINFORMATICS (AGRO-GEOINFORMATICS), 538-542.

Abstract
Crop biomass plays an important role in food security and global carbon cycle. Recently, remote sensing technology has been proven to be an effective tool for biomass estimation with fewer field surveys. In this study, we try to map summer maize biomass using SPOT-5 (Systeme Probatoire d'Observation de la Terre-5) and MODIS (Moderate Resolution Imaging Spectroradiometer) MOD09Q1/A1 data. First, the STARFM (spatial and temporal adaptive reflectance fusion model) algorithm was applied to generate high spatio-temporal resolution NDVI (normalized difference vegetation index) and LSWI (land surface water index) datasets through fusing SPOT-5 and MODIS data. The blended NDVI and LSWI exhibited high correlations with the referenced SPOT-5 NDVI and LSWI (0.72 <= R-NDVI(2) <= 0.76, 0.74 <= R-LSWI(2) <= 0.86). Afterwards, a MLC (maximum likelihood classification) and ISODATA (Iterative Self-Organizing Data Analysis Technique) method was employed to cluster time series NDVI, result shows that 91.3% maize fields were correctly identified and the overall accuracy was 89.8% with a kappa of 0.7957, which was 10-20% higher than the classification of MODIS-NDVI. Finally, NDVI, LSWI and meteorological data were inserted into a light use efficiency model to estimate aboveground biomass, the estimated results that involved blended data correlated well with the measured values (0.66 <= R-2 <= 0.75, 23.5 <= RMSE <= 163.4 g/m(2)), and more accurate than the estimation using MODIS data (0.51 <= R-2 <= 0.66, 27.1 <= RMSE <= 174.4 g/m(2)). In addition, biomass distribution maps were generated based on the crop classification and estimation results. Our research provided a reliable method for the biomass mapping of summer maize with SPOT-5 and MODIS data, which is also very meaningful for the precision agriculture.

DOI:

ISSN:
2334-3168