Monitoring vegetation phenology changes in the Northern Hemisphere with satellite derived NDVIg and NDVI3g data

Theme 3. Meteorological Cooperation
Session Name 3.1 Opportunities and challenges of remote sensing techniques to meet current and future needs for monitoring the Arctic
Datetime Sep 07, 2018 01:55 PM - 02:10 PM (UTC +3)
Presentation Type Oral
Author(s) Fengmei Yao (College of Earth and Planetary Sciences , Universtiy of Chinese Academy of Sciences, China), Jiahua Zhang (Institute of Remote Sensing and Digital Earth (RADI), Chinese Academy of Sciences, China), Qing Chang (Institute of Remote Sensing and Digital Earth (RADI), Chinese Academy of Sciences, China), Huadong Guo (Institute of Remote Sensing and Digital Earth (RADI), Chinese Academy of Sciences, China)
Abstract text

Phenology is a sensitive and critical feature of vegetation and is a good indicator for climate change studies. In this study,  the global inventory modelling and mapping studies (GIMMS) normalized difference vegetation index (NDVI) of the third version (GIMMS NDVI3g) and the previous version (NDVIg) were used to detect the vegetation pheonoly in the Northern Hemisphere. The three most popular satellite start of vegetation growing season (SOS) extraction methods were used. The results showed that the amplitude and the peak values of the NDVI3g are higher than those of the NDVIg curve, which indicated that the SOS derived from the NDVIg (SOSg) was significantly later than that derived from the NDVI3g (SOS3g) based on all the methods for the whole study area . In addition, SOSg and SOS3g both showed an advancing trend during 1982–2006, but that trend was more significant with SOSg than with SOS3g. The difference between SOSg and SOS3g varied among the methods and was partly related to latitude. For the multi-year mean SOS, the difference decreased in the mid- and high-latitude intervals. The GIMMS NDVI3g data-sets seemed more sensitive than the GIMMS NDVIg in detecting information about the ground, and the SOS3g data were better correlated both with the in situ observations and the SOS derived from the Moderate Resolution Imaging Spectroradiometer NDVI. For the northern hemisphere, previous satellite measures (SOS derived from GIMMS NDVIg) may have overestimated the advancing trend of the SOS by an average of 0.032 d/ yr.

 Download to your calendar