R Shiny app: Arctic sea ice concentration and extreme winds
I’ve uploaded a new R Shiny web application. The app examines trends in sea ice concentration and extreme wind events over time. Since both variables, mean monthly sea ice concentration and proportion of days in a month defined as exhibiting extreme winds, are on the same scale ranging from zero to one, I plot both time series together to make it easier to see if and when there is low ice concentration but high winds.
The user has the option of looking at time series lines or bars. By default, annual percentages for a selected month are plotted as lines and decadal means are shown in a barplot. Sea ice concentration uses a composite global climate model (GCM) – a five-model average of individual GCMs – which works well as a robust approach to summarizing monthly means. Individual GCMs are used for wind events since the interest in winds is specifically focused on the extreme upper distribution quantiles and averaging the models would mask much of the extrema. Two RCPs are available for the wind event proportions variable. The GCMs for winds and sea ice are not the same. A different model selection procedure was used for each variable. The GCMs are from CMIP5 (Coupled Model Intercomparison Project 5). The historical and projected daily wind outputs from the original GCMs were first quantile-mapped to a historical reanalysis dataset (ERA-40). More details about the app can be found on the app About tab.
This app is related to two others which focus on wind and temperature events and Arctic sea ice coverage.