Abstract

The present study aims to estimate the spatio-temporal variability of gross primary productivity (GPP) in moist and dry deciduous plant functional types (PFTs) of northwest Himalayan foothills of India using remote sensing-based Temperature-Greenness (TG) model and to study the response of GPP to environmental variables. TG model was implemented in Google Earth Engine platform using Moderate Resolution Imaging Spectroradiometer enhanced vegetation index (MOD13A2) and land surface temperature (MOD11A2) from 2001 to 2018. The mean monthly GPP ranged from 1.80 to 18.57 gCm⁻²day⁻¹ in moist deciduous and from 0.20 to 12.06 gCm⁻²day⁻¹ in dry deciduous PFTs. On site-scale validation with eddy covariance flux tower GPP, the modelled GPP showed R²=0.79 for moist deciduous and R²=0.77 for dry deciduous PFT. Leaf area index showed the highest correlation with the predicted GPP (r = 0.74 for moist and 0.83 for dry deciduous PFTs). The study revealed that TG model could predict the long-term forest GPP with minimum in-situ inputs.

摘要

本研究旨在利用基于遥感的温度-绿度 (TG) 模型估计印度喜马拉雅西北部山麓潮湿和干燥落叶植物功能类型 (PFT) 的总初级生产力 (GPP) 的时空变异性，并研究GPP 对环境变量的响应。TG模型在谷歌地球引擎平台上使用中分辨率成像光谱仪增强植被指数（MOD13A2）和地表温度（MOD11A2）从2001年到2018年。平均月GPP范围从1.80到18.57 gCm ^-2 day ^-1在潮湿落叶从 0.20 到 12.06 gCm ^-2天^-1在干燥的落叶 PFT 中。在使用涡流协方差通量塔 GPP 进行现场规模验证时，模拟的 GPP 显示湿落叶 PFT 的 R ² =0.79 和干燥落叶 PFT 的 R ² =0.77。叶面积指数与预测的 GPP 相关性最高（潮湿落叶 PFT 的 r = 0.74 和干燥落叶 PFT 的 0.83）。研究表明，TG模型可以以最少的原位投入预测长期森林GPP。