The field of environmental change detection and forecasting is rapidly advancing, with a focus on developing practical and effective methods for monitoring and predicting changes in the environment. Recent research has emphasized the importance of leveraging satellite image time series, self-supervised learning, and multi-modal spatiotemporal datasets to improve the accuracy and scalability of change detection and forecasting models. Notably, the use of novel frameworks and architectures, such as hypernetworks and UNet, has shown significant promise in enhancing the performance of global predictive models. Additionally, the integration of external factors, such as weather indicators and climate data, has been found to be crucial for improving the accuracy of forecasting models. The development of visual analytic frameworks and tools, such as FEWSim, has also facilitated the exploration and interpretation of complex simulation results, enabling domain experts to make more informed decisions. Overall, the field is moving towards the development of more robust, scalable, and interpretable models that can effectively support environmental monitoring and decision-making. Noteworthy papers include: Leveraging Satellite Image Time Series for Accurate Extreme Event Detection, which proposes a novel framework for detecting extreme events using satellite image time series, and OPTIMUS, which introduces a self-supervised learning method for detecting persistent changes in satellite images. HydroChronos is also notable for its large-scale, multi-modal spatiotemporal dataset for surface water dynamics forecasting.