2.1 EP data
The EP data was derived by using the growing season integrated annual normalized difference vegetation index (INDVI) data in the circumpolar vegetation dynamics product of global change study (Gonsamo & Chen, 2016). This is the first set of phenological index based plant dynamics product for the entire circumpolar North during 1999–2013, with the spatial resolution unified to 4km×4km (0.03571428°×0.03571428°). This product was generated based on the SPOT VGT data, and its validation showed good agreements with the ground-truth data from many eddy covariance flux tower sites of North American and European deciduous broadleaf, evergreen needle-leaf, and mixed forests, and wetlands (Gonsamo & Chen, 2016). Overall, for the Arctic circumpolar region, this vegetation dynamics product is an improvement to the often-used global MODIS combined land cover dynamics MCD12Q2 product (Gonsamo & Chen, 2016).
Specifically for the derivations of the EP parameters for the lower-EP (LEP), higher-EP (HEP), and all-cases-EP (AEP) NH ecosystems, the average of the lowest 30%, the average of the highest 30%, and the average of the INDVIs for the matrix of 14×14 geo-grids corresponding to any a geo-grid (0.5°×0.5°) in the used SP data (Chen et al.,2016) were calculated by following Equation (1), (2), and (3), respectively.
(1)
(2)
(3)
where denotes the lowest 30%, indicates the highest 30%, is the considered year, and is the number of the geo-grids satisfying the definitions of the related EP parameters for any a 14×14 geo-grid matrix. The theoretical foundation of applying such vegetation indices for characterizing macro-scale vegetation productivity has already been validated in a plenty of previous researches (Wilson & Meyers, 2007; del Castillo et al., 2018). As regards to the setting of 30% in defining the EP of the LEP and HEP ecosystems, such a relatively larger threshold can effectively avoid the impacts of the outliers in some 4km×4km geo-grids (Gonsamo & Chen, 2016) and, hence, can better reflect the productivity of vegetation at the ecosystem level.