The interplay between climate change and societal shifts, particularly the rise and fall of empires, has long been a focus of historical and paleoclimatological research. While the connection is often inferred, concrete examples remain scarce. This study investigates the impact of a significant climatic event, the Late Ming Dynasty Megadrought (LMDMD), on the transition of power from the Ming to the Qing Dynasty in China. The LMDMD is considered a pivotal factor contributing to the Ming Dynasty's collapse, serving as a compelling case study of climate change's influence on socioeconomic and political structures. Previous research, using diverse proxies like stalagmites, lake sediments, and historical documents, has characterized the LMDMD’s duration, intensity, and spatial extent. However, these reconstructions often face limitations in absolute dating and clear physiological significance. Tree-ring data, known for its accurate dating and wide spatial distribution, offers a significant advantage in evaluating the relationship between economic productivity and monsoonal moisture changes. While the existing Monsoon Asia Drought Atlas (MADA) utilizes tree-ring data, there is a notable gap in robust hydroclimatic information from southern China. To address this gap, this study develops a tree-ring chronology from southern China to provide a more comprehensive understanding of the LMDMD and its impact.

Numerous studies have examined the link between the Asian monsoon and ancient Chinese civilizations, emphasizing the impact of monsoon intensity variations on economic activity and social stability. These variations triggered events like peasant uprisings, epidemics, and dynastic changes through disruptions in supply and demand. The LMDMD is widely acknowledged as a key climatic driver in the Ming Dynasty's decline, and numerous climate proxy records have been used to characterize its spatiotemporal characteristics. These studies have implicated external forcings, such as temperature anomalies, reduced solar activity, and internal variability in land-sea heat exchange, as contributing factors. However, limitations in absolute dating and physiological significance in these existing proxies necessitate further investigation using alternative data sources, such as tree rings.

This research utilized tree-ring width (TRW) data from *Keteleeria davidiana* trees in the Daba Mountains of the middle Yangtze River basin. Ninety-five cores were collected from 43 trees, covering a wide geographical area. The ring widths were measured using CooRecorder 9.4 software, and dating accuracy was verified with COFECHA. The regional chronology was developed using ARSTAN, employing the Friedman super smoother for growth trend fitting and a double-weighted robust mean for creating the standard chronology. Variance stabilization was performed using the method by Osborn et al. The chronology’s robustness was confirmed using the expressed population signal (EPS) and inter-series correlation (Rbar). A Pearson correlation analysis examined the relationship between TRW and climatic variables (temperature, precipitation, scPDSI) from nearby weather stations and gridded datasets. The analysis considered both current and previous growing seasons' climate variability. Seven different models (Linear Regression, Neural Networks, Support Vector Machine, Linear Support Vector Machine, K-Nearest Neighbor, Xtreme Gradient Boosting, and Random Forest) were used to reconstruct the April-November self-calibrating Palmer Drought Severity Index (scPDSI) from the TRW chronology. Model performance was evaluated using eight statistical parameters, and an ensemble average of all models was used for the final reconstruction. The reconstruction was then compared with other independent climate records (from stalagmites, lake sediments, and historical documents) to assess the spatiotemporal consistency of the LMDMD across the East Asian monsoon region. Spectral analysis (multi-taper method), ensemble empirical mode decomposition (EEMD), and sliding correlation analysis were used to identify dominant periodicities and correlations with large-scale climate indices (PDO, ENSO) and solar activity and volcanic eruptions. Superposed epoch analysis (SEA) was also used to examine the impact of these external factors on drought.

The *Keteleeria davidiana* tree-ring chronology successfully reconstructed April-November scPDSI for the past 464 years. The reconstruction revealed distinct periods of drought and pluvial, with the LMDMD (1625–1644 CE) exceeding the 100-year return period. Comparison with other proxy records confirmed the consistent spatial structure of the weak monsoon events during the late Ming period across northern and southern China. Although the onset, duration, and magnitude varied regionally, all records indicated a significant weakening of the monsoon during this period. Historical documents corroborate the severe drought, including reductions in crop yields, locust plagues, epidemics, and social unrest that exacerbated the decline of the Ming Dynasty. Spectral analysis of the scPDSI revealed dominant interannual (3–4 years) and decadal (10–30 years) cycles. Correlation analysis showed relationships between drought reconstructions and sea surface temperature (SST) in the western and northern Pacific. The LMDMD coincided with a cold phase of the PDO, a warm phase of ENSO, weakened solar activity (Maunder Minimum), and an increased frequency of volcanic eruptions in the Asia-Pacific region. Superposed epoch analysis revealed that extreme PDO and ENSO phases, along with major volcanic eruptions, significantly influenced scPDSI variations, although relationships were not universally consistent. This suggests that the LMDMD resulted from a combination of internal (PDO, ENSO) and external (solar activity, volcanic eruptions) forcing factors.

The findings of this study demonstrate that the LMDMD was a widespread and impactful event that significantly impacted the Ming Dynasty's stability. The combination of detailed tree-ring data from a previously under-represented region (southern China), and the integration of multiple proxy records and climate indices, provides a comprehensive understanding of the spatial and temporal dynamics of this megadrought. The demonstrated harmonized response to the weakened monsoon across both northern and southern China highlights the interconnectedness of the East Asian monsoon system. The identification of multiple driving forces, including both internal climate variability and external forcings, underscores the complexity of drought mechanisms and the need for multifaceted approaches to understand and predict such events. The study highlights the profound impact of climate change on society and provides valuable insights for developing strategies to mitigate future drought risks in the region.

This study provides a robust, 464-year reconstruction of water balance changes in the middle Yangtze River, confirming the widespread impact of the Late Ming Dynasty megadrought. The integrated analysis of tree-ring data, other climate proxies, and climate indices emphasizes the importance of both internal climate variability and external forcings in driving drought events. This research offers crucial historical context for managing future water resources, particularly given the anticipated impacts of climate change on the region. Future research could explore the detailed mechanisms linking specific climate indices to regional drought patterns and further refine the understanding of the complex climate-societal interactions during the Late Ming Dynasty.

While this study significantly enhances our understanding of the LMDMD, several limitations should be acknowledged. The tree-ring chronology relies on a specific species (*Keteleeria davidiana*) with a limited geographical distribution, potentially influencing the representativeness of the reconstruction. The statistical models used have inherent uncertainties, and the ensemble approach may not fully capture all climate signals. Further studies using alternative proxy records and a broader range of tree species could improve the accuracy and spatial coverage of the reconstruction. Finally, the precise mechanisms linking large-scale climate indices to regional drought are still being actively researched, and more investigation is needed to fully clarify these connections.