Remotely Evaluating the Seasonality of Gross Primary Production at High Latitudes

• Main Author: Cheng, Rui
• Format: Thesis
• Language: English
• Published: 2023
• Subjects: Arctic; Arctic Greening; Climate change
• Online Access: https://thesis.library.caltech.edu/14999/; https://thesis.library.caltech.edu/14999/1/Cheng_Rui_2023_thesis%20%281%29.pdf; https://resolver.caltech.edu/CaltechTHESIS:08112022-214313625

A warming trend larger than the global average is changing high-latitude terrestrial ecosystems. The impact of climate change at high latitudes is especially notable on the seasonality of vegetation photosynthesis, such as the Arctic greening, lengthened growing season, and increased peak production in the growing season. As a critical component of the global carbon cycle and land carbon sink, continuously monitoring the seasonal trajectory of ecosystem-level photosynthesis, Gross Primary Production (GPP), is much needed to better understand the climate change impacts and the sensitivity of high-latitude plant communities under global climate change. GPP has been estimated from both ground and space. However, sparsely distributed ground-level measurements are not representative of heterogeneous land cover and complex terrain in high latitudes. Remote sensing techniques provide extensive spatial coverage for comparing GPP at the regional scale. In this thesis, I carefully examine the advances in remote sensing for monitoring GPP at high latitudes, including using hyperspectral reflectance and Solar-Induced chlorophyll Fluorescence (SIF). We show that reflectance near 531 nm can track the seasonality of Light Use Efficiency (LUE), complementing conventional normalized difference vegetation index which is only a proxy of Absorbed Photosynthetic Active Radiation (APAR). Tracking both LUE and APAR is critical for improving GPP estimation, especially in evergreen forests with photosynthetic phenology but sustained canopy color -- a typical land cover type at high latitudes. Satellite-measured SIF can also track both LUE and APAR. Here, it is shown that the empirical model predicting GPP using SIF is land cover dependent. The presence of snow and surface, heterogeneous land cover, and complex terrains in the high latitudes further complicate the interpretation of the SIF-GPP relationship. To improve the accuracy of interpreting SIF in complex terrain, a geometric model is developed to account for variations in APAR on ...