Obituary: Anders Persson

David Richardson


Anders Persson
Anders Persson in front of the screens at ECMWF (June 2013).

Anders worked in the Meteorological Operations section during the 1990s, a key decade which saw the introduction of medium-range ensemble forecasts in 1992 and the start of the evolution of probabilistic forecasting to the central role it plays today.

Understanding was important to Anders, and he raised many challenging questions that helped to clarify the ensemble approach. He became a great advocate of the benefits of ensembles and put a lot of effort into communicating these to forecast users at every opportunity. This culminated in the development of the ECMWF Forecast User Guide, which presented clearly and in detail the practical approaches by which users could take advantage of the ensemble and integrate this into their work.

Another major component of Anders’ work at ECMWF was the day-to-day operational monitoring of the performance of the forecasting system, and the investigation of cases of poor performance and systematic errors. Anders pioneered the use of several diagnostic approaches, including error tracking to try to identify the sources of forecast error. By always questioning forecast performance and actively following up with staff across the Centre, he helped to ensure that the ECMWF forecasting system continued to improve. Again, much of Anders’ insight into these areas was captured in the Forecast User Guide.

After leaving ECMWF in 2001, Anders always maintained a strong link to the Centre, visiting on many occasions and always keeping a keen eye on forecast performance. During a visit to ECMWF in 2010–11 to prepare a new edition of the Forecast User Guide, Anders was interviewed for the ECMWF Newsletter. This conversation provides a fitting testimony in his own words to all his experience over the years at the Centre:

Anders has had a lasting influence at ECMWF and will be well remembered also by the many users of ECMWF forecasts who attended training courses and workshops at ECMWF, where he was always passionate about promoting the use of ensemble forecasts. Anders died on 23 January 2021.


​I was really saddened and shocked, like all of us, by this news. Anders was still so active and passionate the last time I saw him not that long ago. I have very fond memories of Anders since my first stay at ECMWF in the 1990s. He never lost touch afterwards and had so many intellectually challenging discussions with many of us. He will be sadly missed but not forgotten! (ECMWF Director-General Florence Rabier) 

Many of us will be saddened to hear of the death of Anders Persson. Anders chose a career in meteorology in 1964 in the mistaken belief that meteorology was a simpler science than quantum mechanics or relativity theory. He followed at close range – and with great interest – the spectacular development of numerical weather prediction.​ (Austin Woods)

Anders was a wonderful colleague to share ideas with. He was very widely read and would engage pretty much on any topic you wanted to. Why, he would argue, do most of the ensemble members have a higher root-mean-square error than the unperturbed control? Didn’t that mean something was wrong? I will miss our intellectual debates.​ (Tim Palmer)

​I am really very sorry to hear this sad news. I worked a lot with Anders. I liked to call him ‘the ensemble conscience’, since he was always asking very appropriate and challenging questions about it. Thanks to his questions and engaging discussions, I think we did indeed check more carefully the language we were using. (Roberto Buizza)

I met Anders for the first time around 1980 when he came to spend a few weeks in Paris to see how French forecasters were working. At ECMWF we often had long discussions about the meteorological situation of the day. After I left ECMWF, he contacted me several times on scientific-historical subjects, and I learnt a lot from the subsequent exchanges with him. (Jean Pailleux)

I had the difficult task, and the honour, to replace Anders as a synoptic meteorological analyst at ECMWF. I did it with the same enthusiasm and fascination that he transmitted to me during my training. He is still in the Italian forecasting community’s memory after the many exciting lectures he used to give at the MeteoTrentino courses. (Federico Grazzini)