Publications

Zhang, BY; Liu, XN; Liu, ML; Meng, YY (2019). Detection of Rice Phenological Variations under Heavy Metal Stress by Means of Blended Landsat and MODIS Image Time Series. REMOTE SENSING, 11(1), 13.

Abstract
Monitoring phenological changes of crops through remote sensing methods is becoming a new perspective in assessing heavy metal contamination in agricultural farmlands. This paper proposes a method that combines the normalized difference vegetation index (NDVI) and the normalized difference water index (NDWI) to detect heavy metal stress-induced variations in satellite-derived rice phenology. First, we applied the enhanced spatial and temporal adaptive reflectance fusion model to obtain the NDVI and NDWI time series for the NDVI-NDWI phase-space construction. Then, six specific rice phenometrics were derived from the NDVI and the phase-space, respectively. Last, we introduced a relative phenophase index (RPI), which characterizes the relative change of the phenometrics to identify the rice paddies under heavy metal stress. The results indicated that satellite-derived rice phenometrics are generally influenced by human and natural factors (e.g., transplanting date, air temperature, and solar radiation), while the RPI showed weak correlations with all of these variables. In the determination of heavy metal stress, the NDVI- and phase-space-based RPIs of unstressed rice both show significantly (p < 0.001) higher values than those of stressed rice, while the phase-space-based RPI shows more apparent statistical difference between the stressed and unstressed rice compared to the NDVI-based one. Our work proved the capability of the phase-space-based method as well as the RPI in the discrimination of regional heavy metal pollution in rice fields.

DOI:
10.3390/rs11010013

ISSN: