Hazard Monitoring at KNMI Netherlands By Frank Kroonenberg

Hazard Monitoring at KNMI (Netherlands) By Frank Kroonenberg (crisis coordinator/senior meteorologist at KNMI)

Risks from Meteorology and Vulnerability Small country approx 42. 000 km 2 with high density of people 495 persons/km 2 Initial impact is strongly determined by dynamical initial vulnerabilities such as: • Density of population and economical activities • Working or none working days on the calendar • Traffic rush hours • Outside festivities 2 9/9/2020

Hitlist highest impact parameters 1. Meteorological driven hydrological events (River and Coastal floodings) We are very well protected against this main attack from the weather. High dikes with flooding risk once in 10. 000 years for our coastal defence and once in 1. 250 years for rivers. Also man made barrières to protect weak spots on our coastal line 3 9/9/2020

4 9/9/2020

5 9/9/2020

6 9/9/2020

7 9/9/2020

Hit list highest impact parameters 1. Meteorological driven hydrological events (River and Coastal floodings) We are very well protected against this main attack from the weather. High dikes with flooding risk once in 10. 000 years for our coastal defence and once in 1. 250 years for rivers 2. Windstorms and wind gusts Synoptical wind storms accurately predictable at proper lead times, Wind gusts from convective systems far more difficult to predict. High impact wind events happen once in the one or two years 8 9/9/2020

9 9/9/2020

Hit list highest impact parameters 1. Meteorological driven hydrological events (River and Coastal floodings) We are very well protected against this main attack from the weather. High dikes with flooding risk once in 10. 000 years for our coastal defence and once in 1. 250 years for rivers 2. Windstorms and wind gusts Synoptical wind storms accurately predictable at proper lead times, Wind gusts from convective systems far more difficult to predict (high impact wind events 0. 5 till 1 per year) 3. Tornadoes are quit rare and not predictable at good lead times. We do not have a Tornado watch or warning system. Return periods for a F 1/F 2 Tornado is once 10 -20 yrs, F 3 once 20 -30 yrs 10 9/9/2020

Tornado Borculo 1925 11 9/9/2020

Hazard monitoring and detecting from obs and by info from CP partners We do not maintain hazard database or perform hazard mapping or analyses on a routine basis Main reason: Next to 4 to 7 cases each year with “relatively large” disruption and few casualties, there are hardly any very high impact events However we do evaluate all orange and red warning cases (4 to 7/yr), not only from a met perspective but combined with impact info from CP partners. This in order to optimise warning thresholds settings. Next to this we use statistical insurance information out of a combination from weather parameter intensities and impact (costs) 12 9/9/2020

13 9/9/2020

Prevention and repression Since dynamic part of vulnerability is that important we set warning thresholds, primarily based on climate return periods, but together with CP partners We involve CP partners to choose initially between an orange or a red warning (KNMI + impact expert partners) We do have an advanced alerting system towards the Hydro institute with 5 days lead times, when EPS + Expert judgement fcst critical storm surge levels on our coast Also during critical hydro risk weather, specially trained KNMI meteorologists work at The Water Management Centre Netherlands (WMCN) together with hydrologists to determine the risks and give advisories to the competent authorities Climate change studies are taken into account for prevention planning on hydro events defence (higher dikes, water storage areas) 14 9/9/2020

KNMI Return Period Analysis for Wind gusts (1 + 0. 5 /yr) 15 9/9/2020

Summary and Conclusions KNMI doesn’t perform natural hazard statistics on a routine basis, by reasons: We have few very real severe impact events happening Most of these events are triggered by weather and high initial vulnerability. So only looking at weather parameter intensities, not taking initial vulnerability (which is really complex to do as NMS) into account, makes us possibly look at wrong data for weather impact analysis What we do: We do perform evaluations and threshold and impact studies together with CP after every event (for orange and red). In order to learn and to adjust thresholds and procedures every few years 16 9/9/2020

17 9/9/2020