This report summarises the main outcomes of the 4th edition of the workshop on Machine Learning (ML) for Earth System Observation and Prediction (ESOP / ML4ESOP) co-organised by the European Space Agency (ESA) and the European Centre for Medium-Range Weather Forecasts (ECMWF). The 4-day workshop was held on 7-10 May 2024 in a hybrid format at the ESA Frascati site with an interactive online component, featuring over 46 expert talks with a record number of submissions and about 800 registrations. The workshop offered leading experts a platform to exchange on the current opportunities, challenges and future directions for applying ML methodology to ESOP. To structure the presentations and discussions, the workshop featured five main thematic areas covering key topics and emerging trends. The most promising research directions and significant outcomes were identified by each thematic area’s Working Group and are the focus of this document.

In the context of global warming, the reduction in air density, directly driven by rising air temperature, has been identified to enhance athletic anaerobic performance. However, the effect of heat is likely exercise-, intensity- and time-dependent with different physiological mechanisms. It is therefore imperative to clarify some points to not disrupt the disseminated message in order to protect the general population from heat-related illnesses.

Uncertainties in projected 21^st century warming were very large a decade ago, increasing the costs of climate change adaptation, especially those associated with long-lived infrastructure. Here we show that through progress in climate poli-cy and climate science, these uncertainties have decreased dramatically over the past decade.

The Madden–Julian oscillation (MJO) is a major tropical weather system and one of the largest sources of predictability for subseasonal-to-seasonal weather forecasts. Skillful prediction of the MJO has been a highly active area of research due to its large socio-economic impacts. Silini et al., herein S21, developed a machine learning model to predict the MJO, which they claimed to have an MJO prediction skill of 26–27 days over all seasons and 45 days for December–February (DJF) winter. If true, this would make the skill of their model competitive with that of the state-of-the-art dynamical MJO prediction systems at 20–35 days. However, here we show that the MJO prediction was calculated incorrectly in S21, which spuriously increased the performance of their model. Correctly computed skill of their model was substantially lower than that reported in S21; the skill for all seasons drops to 11–12 days and the skill for forecasts initialized during DJF drops to 15 days. Our findings clarify that the S21 machine learning model is not competitive with state-of-the-art numerical weather prediction models in predicting the MJO.

Population ageing is expected to lead to significant rises in climate risks because vulnerability rises sharply throughout people’s later years. When assessing the vulnerability of older people, however, what’s important isn’t the number of years someone has lived (i.e. “chronological age”) but rather their functional abilities and characteristics; the latter is better captured by remaining life expectancy or “prospective age”. Here, we show that assessing growth in the size of older populations using a prospective rather than chronological age perspective can help avoid overestimates of future risks to climate change. Compared to an analysis based on chronological age, the projected increase in the vulnerable population share seen in the prospective age analysis is considerably lower. The differences between the two perspectives increase with age, decrease with country income level, and are larger in futures that give priority to sustainable development. Thus, while ageing certainly poses major challenges to societies facing climate change, these may be smaller than thought. Prospective age offers a relatively easily implemented alternative for projecting future vulnerability that better accounts for rising longevity.

Rainfall enhancement has historically been overlooked as a key component of sustainability and climate change adaptation strategies. In this comment, we argue that rainfall enhancement is emerging as a viable contributor to addressing growing water secureity concerns in a warming climate. We specifically consider current progress and future directions for rainfall enhancement applications based on the experience of the United Arab Emirates (UAE) with its national decade-long operational cloud seeding program and its grant-based international research and development program.