Publications

Liu, F; Wang, XC; Wang, CK (2019). Autumn phenology of a temperate deciduous forest: Validation of remote sensing approach with decadal leaf-litterfall measurements. AGRICULTURAL AND FOREST METEOROLOGY, 279, UNSP 107758.

Abstract
Autumn (i.e., leaf-fall) phenology plays an important role in regulating the canopy duration and is often monitored using near-surface and satellite remote sensing techniques, but the measurements have rarely been validated by the ground observation. The objectives of this study were to: (1) evaluate the performance of radiometer-based broadband vegetation index (VIB) and Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation index (VIM) in monitoring the autumn phenology of a Chinese temperate deciduous forest with decadal (2008 - 2018) leaf-litterfall measurements, and (2) explore the feasibility of using VIB to validate the VIM products for monitoring the autumn phenology. We found that the seasonal and interannual trends of both VIB and VIM agreed well with those of the leaf-litterfall (correlation coefficients r > 0.93). The broadband normalized difference vegetation index (NDVIB) best tracked the interannual variation in the end-of-season (EOS) among the six metrics of the VIB investigated, while the MODIS enhanced vegetation index (EVIM) did so among the six metrics of the VIM, with the corresponding determination coefficient (R-2) of 0.66 and 0.44 with the EOS estimated by the leaf-litterfall. The EOS estimated by VIB and VIM was 0 - 11 d earlier than that by the leaf-litterfall. Comparing the six metrics of VIM with the corresponding ones of VIB, the EOS derived from the NDVI had the closest correlation with each other (R-2 = 0.67). Conclusively, our study validated the remote-sensed leaf-fall phenology with decadal ground measurements, and suggested that radiometer and MODIS could effectively track the autumn phenology in temperate deciduous forests and the leaf-litterfall collection could be used as a complementary approach.

DOI:
10.1016/j.agrformet.2019.107758

ISSN:
0168-1923