Integrating solar-induced chlorophyll fluorescence and nighttime temperature to enhance the detection of the start of growing season in evergreen broadleaf forests under monsoon climates

Accurate phenological information is crucial for assessing ecosystem dynamics and carbon budgets. Evergreen broadleaf forests (EBF), as a typical evergreen vegetation type in tropical and subtropical areas, exhibit strong carbon sequestration capacity and play important roles in maintaining biodiversity and regulating climate. However, remote sensing monitoring methods based on a single index remain challenging for achieving high-precision phenological extraction in EBF. Yunnan Province, a key carbon sink in southwestern China, has a 55.04% forest coverage, with EBF dominating nearly half of this area and driving biodiversity. Consequently, this study focuses on phenological extraction in EBF of Yunnan. Firstly, we evaluated the performance of several existing phenological extraction models for evergreen vegetation. Secondly, based on existing models, we propose a novel framework for extracting start of growing season (SOS) in EBF. The framework integrates solar-induced chlorophyll fluorescence (SIF) and temperature variables, combining multiple curve-fitting methods and phenological extraction approaches. Finally, by comparing with ground observations, this study validated the capability of the framework to extract SOS in EBF, determined the optimal extraction scheme, and analyzed the spatiotemporal patterns and trends of SOS in EBF in Yunnan Province from 2004 to 2013. Results show (1) Accuracy evaluation metrics ( R 2 , RMSE, and P -value) indicate that existing models generally yield low-precision SOS extraction for EBF in the study area. (2) Compared to existing models, the proposed framework improves R 2 by 0.4 (from 0.35 to 0.75) and reduces RMSE by 5.86 days (from 15.19 to 9.33 days). (3) The SOS for EBF was mainly concentrated 90–110 days of the year, showing an overall advancing trend. The methodological framework presented in this study can serve as a valuable reference for the future development of higher precision phenological models for EBF, thereby contributing to more accurate assessments of ecosystem dynamics and carbon uptake.