The research team innovates forest phenological monitoring technology: SIF remote sensing data

Forests play a crucial role in the global carbon cycle, especially in climate regulation and stability. However, accurate monitoring of temporal changes in forest phenology, especially photosynthesis, has been a challenge in the field of ecology. Traditional vegetation indices, such as normalized Differential Vegetation Index (NDVI), have been widely used in many occasions, but these indices only reflect the "greenness" of vegetation. In dense vegetation ecosystems such as forests, signal saturation often exists, which makes it difficult for NDVI to accurately capture forest photosynthetic phenology. The SIF (Solar induced fluorescence) remote sensing data obtained through cooperative research can more directly reflect the dynamics of plant photosynthesis, but the existing SIF data still have some challenges in spatial resolution and time continuity. To overcome these problems, high spatio-temporal resolution SIF data needs to be reconstructed to extract photosynthetic phenologies of forests.

First, on February 20, 2025, a research team from Wuhan University and Emory University published a groundbreaking study in the Journal of Remote Sensing, proposing a maximum vertical distance (CBPD) method for estimating the phenological change point of forest photosynthesis. This innovative method uses SIF data to determine the beginning and end of the growing season more accurately than traditional methods, which provides an important new tool for ecological research and climate change research, and is expected to revolutionize the field of forest phenological monitoring.

Secondly, one of the highlights of the research is the application of CBPD method, whose accuracy is significantly better than traditional techniques such as double logic (DL), first derivative (FOD) and dynamic threshold (DT). The CBPD method showed significant results advantages over traditional methods: root mean square error (RMSE) was reduced by 0.04 to 14.04 days, Pearson correlation coefficient (R) was improved by 0.03 to 0.30, and consistency index (AI) was improved by 0.34 to 21.52. These significant improvements indicate that the CBPD method has made a qualitative leap in both accuracy and stability.

At the same time, the study also revealed that the main driver of spring phenology change is temperature, while the fall phenology is more affected by radiation. The finding highlights the profound effects of climate change on forest growth patterns. To verify the effectiveness of the CBPD method, the team used the global OCO-2 SIF (GOSIF) dataset, which combines machine learning, MODIS data, and meteorological reanalysis to provide high spatio-temporal resolution photosynthesis information. The research team verified the reliability of the CBPD method by comparing data from 38 eddy covariance flux tower stations across North America, spanning 2001 to 2020, with daily total primary productivity (GPP) and net ecosystem exchange (NEE) data from flux towers. The results show that CBPD is superior to traditional techniques in complex climatic conditions and diverse vegetation types.

In addition, the lead researcher of the study said: "This work provides a completely new perspective for forest phenological monitoring. CBPD not only improves the accuracy of phenological parameter extraction, but also provides valuable insights into how forests respond to climate change. We will further expand the application of this method to cover more ecosystems and vegetation types, unlocking its greater potential."

Finally, the CBPD method has a broad potential application prospect, especially in the management and protection of forest ecosystems, and can provide a solid scientific basis for relevant decisions. In addition, with the continuous improvement of SIF data resolution, CBPD method is expected to be applied in more ecosystems, bringing new breakthroughs for global climate change monitoring and ecological environment research. With the continuous advancement of technology, this method not only provides a new research perspective for ecology, but also lays a more efficient way for global environmental monitoring.

In summary, the CBPD method marks a leap in forest phenological monitoring technology. Through high spatio-temporal resolution SIF data, combined with innovative algorithms, the monitoring of forest photosynthesis phenology will become more accurate, providing more effective support for climate change research and ecological protection. In the future, with the continuous improvement of data and technology, forest ecosystems and their role in climate change will be more clearly presented to us.