ALBERT

Description: Heatwaves have disastrous impacts on human health, ecosystems, and agriculture. Heatwave impacts can be mitigated by early warnings, which were found to effectively reduce the risk from temperature extremes on the sub-seasonal prediction timescale. The quality of heatwave warnings is bound to the predictability of heatwaves, which is mainly investigated in terms of the prediction potential of heatwave intensity. The prediction of heatwave features such as heatwave onset and duration would benefit impact prediction and early warnings. For example, higher mortality risk was previously identified for heatwaves that were more intense, longer, or occurred with earlier timing in the summer. We assess the predictability of heatwave onset, duration, and intensity for the large-scale European heatwaves during 1998-2017. The predictability of heatwave characteristics is evaluated for the lead times of 1—3 weeks in the ECMWF forecast system. The highest predictability in heatwave onset and duration is found over Northern Europe, while Western Europe has the lowest bias in heatwave intensity. Furthermore, we identify the most predictable and least predictable heatwaves over Europe. The most predictable events include the Russian heatwaves of 2010 and 2017, and the 2012 event over Eastern Europe. Some of the least predictable events include the events of 2016 over Russia and of 2017 over Western Europe. The identification of the most and least predictable heatwaves sets the basis for a further investigation of the causes for differences in heatwave predictability over Europe.

Description: Socioeconomic livelihoods in the Horn of Africa (HA) are highly dependent on seasonal rainfall, which occurs during two main seasons: October-November-December (OND) and March-April-May (MAM). During the two last decades the HA region has been affected by severe and prolonged droughts, leading to acute food insecurity, shortage of drinking water, and increasing risk of disease. Sub-seasonal drought prediction over the HA, from two weeks to two months, is therefore crucial for decision making and early warnings across several sectors. The sub-seasonal prediction of high and low precipitation extremes (PEs) by dynamical forecast systems is challenging for both rainy seasons, but there may be potential for extending the current prediction timescale based on remote drivers. To investigate the sub-seasonal predictability of PEs during the OND season we build a Long Short-Term Memory (LSTM) Neural Network predicting biweekly precipitation tercile categories over the HA region. The LSTM is trained on observational and reanalysis data during the period 1981—2020 and provides predictions with lead times of one week to one month. The results show that floods can be more skillfully predicted than droughts for all lead times. Moreover, we use explainable AI methods to explore the contribution of remote drivers to the predictions and potential sub-seasonal forecast opportunities for PEs. Preliminary results show that the sea surface temperature over the tropical Pacific is important for the LSTM prediction, but further investigation is needed to determine more factors affecting the prediction skill for PEs over the HA region.