Key weather extremes affecting potato production in The Netherlands

The possible impact of climate change on frequency and severity of weather extremes is hotly debated among climate scientists. Weather extremes can have a significant impact on agricultural production, but their effect is often unclear; this due to interaction with other factors that affect yield and due to lack of precise definitions of relevant weather extremes. We show that an empirical analysis of historical yields can help to identifying such rare, high impact climate events. A reconstructed time series of ware potato production in Flevoland (The Netherlands) over the last 60 years (1951–2010) enabled us to identify the two main yield affecting weather extremes. In around 10% of the years yield anomalies were larger than −20%. We found that these anomalies could be explained from two weather extremes (and no other), namely a wet start of the growing season and wet end of the growing season. We derived quantitative, meteorological definitions of these extremes. Climate change scenarios for 2050 show either no change or increased frequency of the two extremes. We demonstrate there is large uncertainty about past and future frequencies of the extremes, caused by a lack of sufficiently long historical weather records and uncertainties in climate change projections on precipitation. The approach to identify weather extremes presented here is generally applicable and shows the importance of long term crop and weather observations for investigating key climatic risks to production.

The figure shows five time series of fresh potato yields. The arrows point to years with the largest yield anomalies and lists the cause of each anomaly.Figure optionsDownload as PowerPoint slideHighlights
► Presenting an approach to identifying agronomically relevant weather extremes.
► Historical reconstruction of 50 years of potato yield data in The Netherlands.
► Key weather extremes identified are wet start and wet end of growing season.
► Past and future frequencies of the extremes are uncertain.
► Previous climate change impact studies did not include these extremes.