Final Activity Report Summary - GAP (Geostatistical analysis of precipitation data)
Spatial regionalisation of rainfall serves an important role in climate science, particularly in terms of visualising spatial patterns and providing rainfall estimates at locations where there currently are no observations. Furthermore, regular estimates of rainfall in space and time are often required in ecosystem and hydrological modelling or used for validation exercises of numerical weather or climate models or climate models.
Commonly, point measurements of rainfall are regionalised in space using some form of spatial interpolation. Various techniques are available and depending on the purpose of the interpolated product, different methods may be more or less suitable. The purpose of GAP (Geostatistical Analysis of Precipitation data) was to investigate and implement geostatistical tools for long term analysis of rainfall for the United Kingdom and the Iberian Peninsula. Geostatistics is an applied branch of spatial statistics, developed within the earth science community but now widely used within other environmental sciences. However, many of the tools developed within the framework of geostatistics are yet not commonly used within climate science despite the regular need for spatial regionalisation of meteorological variables. GAP provided an opportunity to explore the use of geostatistics to estimate rainfall for unsampled locations in space and time.
The method chosen to regionalise rainfall for the study areas builds on the concept of stochastic simulation. This methodology represents a development within geostatistics that has become increasingly popular in earth sciences during the previous decade. The way in which is was implemented in GAP, presented two important qualities for spatial regionalisation of rainfall data, namely: (i) the possibility to make an assessment of uncertainty in each rainfall estimate due to the estimation process and (ii) the potential to draw upon information in the time domain in addition to the more commonly used space domain. These aspects of regionalisation are particularly important in regions were data are sparse and estimates can be very uncertain. Furthermore, the use of information in time showed to be important when analysing daily rainfall from the Iberian Peninsula, where spatial dependence in rainfall is very low.
Commonly, point measurements of rainfall are regionalised in space using some form of spatial interpolation. Various techniques are available and depending on the purpose of the interpolated product, different methods may be more or less suitable. The purpose of GAP (Geostatistical Analysis of Precipitation data) was to investigate and implement geostatistical tools for long term analysis of rainfall for the United Kingdom and the Iberian Peninsula. Geostatistics is an applied branch of spatial statistics, developed within the earth science community but now widely used within other environmental sciences. However, many of the tools developed within the framework of geostatistics are yet not commonly used within climate science despite the regular need for spatial regionalisation of meteorological variables. GAP provided an opportunity to explore the use of geostatistics to estimate rainfall for unsampled locations in space and time.
The method chosen to regionalise rainfall for the study areas builds on the concept of stochastic simulation. This methodology represents a development within geostatistics that has become increasingly popular in earth sciences during the previous decade. The way in which is was implemented in GAP, presented two important qualities for spatial regionalisation of rainfall data, namely: (i) the possibility to make an assessment of uncertainty in each rainfall estimate due to the estimation process and (ii) the potential to draw upon information in the time domain in addition to the more commonly used space domain. These aspects of regionalisation are particularly important in regions were data are sparse and estimates can be very uncertain. Furthermore, the use of information in time showed to be important when analysing daily rainfall from the Iberian Peninsula, where spatial dependence in rainfall is very low.