The goal of this reporting tool was to provide a visual overview of the numerous runs from numerical weather prediction models (NWP) during a tropical storm event. Output (PNG or PDF) is suited for insurance client reporting purposes.

Example report for Major Hurricane Irma issued on 9 September 2017. Tracks from various models (black solid lines), isoatchs of gust wind speed (colour lines) and forecasted storm intensity (markers) are shown here.

Figure 1: Example report for Major Hurricane Irma issued on 9 September 2017. Tracks from various models (black solid lines), isoatchs of gust wind speed (colour lines) and forecasted storm intensity (markers) are shown here.

Information on the uncertainty of the track forecast can be gained from the dispersion of models. Often the forecasts shifts significantly over time. During the Hurricane Irma event, initital forecasts suggested a northerly track, which would have spared Florida from the storm. (see figure 2). Later forecasts correctly anticipated the track hitting Florida from the south.

Example report for Major Hurricane Irma issued on 6 September 2017. Multiple models from multiple model runs (issuance times) are shown together here.

Figure 2: Example report for Major Hurricane Irma issued on 6 September 2017. Multiple models from multiple model runs (issuance times) are shown together here.

Technology

  • R packages: data.table, ggplot2