This guidance has been developed to help producers of official statistics analyse and provide informed communication of the potential effects of weather or climate on their data. If you want to analyse your data to see if they are affected by weather or climate, you will need to know 1. What weather or climate data are available. 2. Appropriate methods and tools for analysis. 3. How to communicate results to users. This guidance is designed to assist producers of official statistics with accessing and using weather and climate data, conducting an analysis and communicating results to users. This guidance is aimed at staff involved in the production or analysis of official statistics who are interested in exploring the use of weather or climate data to help them interpret their own data. Concepts discussed in some sections, such as on time series models, can become technical and some familiarity with concepts such as regression is assumed. This guidance provides information on what weather and climate data are available from the Met Office, describing the structure of the data and how they can be accessed and used for analysis. A flow chart with hyperlinks to sections of the guide below provides steps that can be followed to conduct an analysis, from the planning stages through to communication of results using time series models commonly used in the production of seasonally adjusted official statistics. This guidance covers some points on communicating the results of an analysis that uses weather or climate data, but users are also encouraged to consult the GSS guidance on writing about statistics (https://gss.civilservice.gov.uk/statistics/methodology-quality/methodol…) in particular the section on interpretation. This guidance is not a textbook on time series regression analysis using weather and climate data. It introduces concepts and points users to sources of further information and provides examples of analyses that users should be able to replicate from the text. This guidance can be used in different ways depending on the needs of the user. The flow chart below gives an overview of how to conduct an analysis. The text in the flow chart links to relevant sections of the guidance. While there is an order to the guide it is not written with the intention that users will read it cover to cover. The panel on the right contains hyperlinks to examples of analyses using weather or climate data. These examples contain all the data and code used to create the analyses so that users of the guide can recreate the examples. The retail sales example describes an analysis of the effect of temperature on monthly estimates of clothing sales and finds evidence a ‘switching effect’ where sales are brought forward in a warmer than usual spring period or a colder than usual autumn. The analysis of road accidents explores the effect of temperature on monthly road accident statistics, finding evidence of increased accidents during warmer than usual periods for the time of year. The ambulance response time analysis explores potential weather effects on the proportion of ambulance call outs that met the eight minute target in Cardiff. The models tested found no evidence of a weather effect. (Author/publisher)
Samenvatting