Philip Stier

Abstract:

Realistically representing deep atmospheric convection is important for accurate numerical weather and climate simulations. However, parameterizing where and when deep convection occurs (“triggering”) is a well-known source of model uncertainty. Most triggers parameterize convection deterministically, without considering the uncertainty in the convective state as a stochastic process. In this study, we develop a machine learning model, a random forest, that predicts the probability of deep convection, and then apply clustering of SHAP values, an explainable machine learning method, to characterize the uncertainty of convective events. The model uses observed large-scale atmospheric variables from the Atmospheric Radiation Measurement constrained variational analysis dataset over the Southern Great Plains, US. The analysis of feature importance shows which mechanisms driving convection are most important, with large-scale vertical velocity providing the highest predictive power for more certain, or easier to predict, convective events, followed by the dynamic generation rate of dilute convective available potential energy. Predictions of uncertain, or harder to predict, convective events instead rely more on other features such as precipitable water or low-level temperature. The model outperforms conventional convective triggers. This suggests that probabilistic machine learning models can be used as stochastic parameterizations to improve the occurrence of convection in weather and climate models in the future.

Abstract:

<jats:title>ABSTRACT</jats:title><jats:p>Changes in “blue water”, which is the total supply of fresh water available for human extraction over land, are quite closely related to changes in runoff or equivalently precipitation minus evaporation, . This article examines how climate change‐driven recent past and future changes in the regional water cycle relate to blue water availability and changes in human blue water demand. Although at the largest scales theoretical and numerical model predictions are in broad agreement with observations, at continental scales and below models predict large ranges of possible future and runoff especially at the scale of individual river catchments and for shorter timescale subseasonal floods and droughts. Nevertheless, it is expected that the occurrence and severity of floods will increase and that of droughts may increase, possibly compounded by human‐driven non‐climatic changes such as changes in land use, dam water impoundment, irrigation and extraction of groundwater. Contemporary assessments predict that increases in 21st century human water extraction in many highly‐populated regions are unlikely to be sustainable given projections of future . To reduce uncertainty in future predictions, there is an urgent need to improve modeling of atmospheric, land surface and human processes and how these components are coupled. This should be supported by maintaining the observing network and expanding it to improve measurements of land surface, oceanic and atmospheric variables. This includes the development of satellite observations stable over multiple decades and suitable for building reanalysis datasets appropriate for model evaluation.</jats:p>

Abstract:

Over the past decades, there has been an explosion in the amount of available Earth observation (EO) data. The unprecedented coverage of Earth’s surface and atmosphere by satellite imagery has resulted in large volumes of data that must be transmitted to ground stations, stored in data centers, and distributed to end users. Modern Earth system models (ESMs) face similar challenges, operating at high spatial and temporal resolutions, producing petabytes of data per simulated day. Data compression has gained relevance over the past decade, with neural compression (NC) emerging from deep learning and information theory, making EO data and ESM outputs ideal candidates because of their abundance of unlabeled data.

In this review, we outline recent developments in NC applied to geospatial data. We introduce the fundamental concepts of NC, including seminal works in its traditional applications to image and video compression domains with a focus on lossy compression. We discuss the unique characteristics of EO and ESM data, contrasting them with “natural images,” and we explain the additional challenges and opportunities they present. Additionally, we review current applications of NC across various EO modalities and explore the limited efforts in ESM compression to date. The advent of self-supervised learning (SSL) and foundation models (FMs) has advanced methods to efficiently distill representations from vast amounts of unlabeled data. We connect these developments to NC for EO, highlighting the similarities between the two fields and elaborate on the potential of transferring compressed feature representations for machine-to-machine communication. Based on insights drawn from this review, we devise future directions relevant to applications in EO and ESMs.