In recent decades, flash drought events have frequently occurred in the humid regions of southern China. Due to the sudden onset and rapid intensification of these droughts, they often cause severe damage to vegetation photosynthesis. However, our understanding of the spatiotemporal evolution characteristics of flash droughts across different vegetation types, as well as the response regularity of photosynthesis to flash droughts, especially early responses, remains limited. This study analyzes the spatiotemporal evolution characteristics of flash droughts for different vegetation types in the Middle and Lower Reaches of the Yangtze River Basin from 2000 to 2023. It uses solar-induced chlorophyll fluorescence (SIF) and fluorescence yield (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\upvarphi\:}}_{\text{F}\:}$$\end{document}) to explore the response regularity of vegetation photosynthesis to flash droughts, with a systematic analysis of the 2013 flash drought event. The results show that, over the past 24 years, the frequency of flash droughts for different vegetation types in the Middle and Lower Reaches of the Yangtze River Basin has decreased, but the total duration has increased, with forests experiencing the highest frequency of flash droughts, while cropland experiences the least. Cropland photosynthesis is the most sensitive to flash drought, showing an early response 8-16 days after the onset and reaching a negative anomaly between 24 and 32 days. Forests mainly show an early response between 16 and 24 days and a negative anomaly response between 32 and 40 days. During the 2013 flash drought, cropland showed an early response on the 10th day after the onset and a negative anomaly on the 26th day, while forest responses were later, with early responses on the 20th day and negative anomalies on the 36th day. These results align with long-term statistical data. This study contributes to a deeper understanding of vegetation photosynthesis response regularity to flash droughts and provides insights for developing effective flash drought management strategies.

The boreal forest is a major contributor to the global climate system, therefore, reducing uncertainties in how the forest will respond to a changing climate is critical. One source of uncertainty is the timing and drivers of the spring transition. Remote sensing can provide important information on this transition, but persistent foliage greenness, seasonal snow cover, and a high prevalence of mixed forest stands (both deciduous and evergreen species) complicate interpretation of these signals. We collected tower-based remotely sensed data (reflectance-based vegetation indices and Solar-Induced Chlorophyll Fluorescence [SIF]), stem radius measurements, gross primary productivity, and environmental conditions in a boreal mixed forest stand. Evaluation of this data set shows a two-phased spring transition. The first phase is the reactivation of photosynthesis and transpiration in evergreens, marked by an increase in relative SIF, and is triggered by thawed stems, warm air temperatures, and increased available soil moisture. The second phase is a reduction in bulk photoprotective pigments in evergreens, marked by an increase in the Chlorophyll-Carotenoid Index. Deciduous leaf-out occurs during this phase, marked by an increase in all remotely sensed metrics. The second phase is controlled by soil thaw. Our results demonstrate that remote sensing metrics can be used to detect specific physiological changes in boreal tree species during the spring transition. The two-phased transition explains inconsistencies in remote sensing estimates of the timing and drivers of spring recovery. Our results imply that satellite-based observations will improve by using a combination of vegetation indices and SIF, along with species distribution information. Plain Language Summary The boreal forest is one of the most sensitive regions on the planet to climate change, yet its sensitivity remains poorly understood. In particular, the timing and drivers of the spring transition, as the forest changes from a winter adapted state to a summer adapted state, carry significant uncertainties. Remote sensing metrics can be used to characterize the spring transition, but their interpretation is complicated by persistent greenness, frequent snow cover, and a high prevalence of forests containing both deciduous and evergreen species. We collected tower-based remotely sensed metrics, stem radius, and carbon uptake measurements and show that the spring transition occurs in two distinct phases. The first phase is a reactivation of photosynthesis in evergreens and is triggered by thawed stems, warm air temperature, and moist soil. The second phase is a change in evergreen photoprotective pigment levels and the leaf-out of deciduous species. It is triggered by soil thaw. Both phases were detected with different remote sensing metrics that depended on species type. Our results illustrate how satellite measurements could be improved to capture the spring transition over diverse landscapes and what environmental factors control the spring transition.