To the list of courses || To: RASA2018 || Estonian

Spatial Correlation and Regression

Correlation is mutual relationship between numerical variables. Spatial correlation differs from non-spatial by including location parameters (usually coordinates). Usually, the dependence of correlation from distance is studied or the weight of observation pairs is set dependent on their distance. Regression is dependence of a numerical variable on explanatory numerical variables.

Both correlation and regression can be calculated locally at each location or grid node. The strength and significance and other parameters of relationshiops can be presented as cartograms.

Not answered
The answer needs to be updated
ID P Question or exercise
3931 2 The example data in the SDC Local Correlation module are the long term mean durability of snow cover and long term mean annual precipitation amount in Estonian meteorological stations.
  1. In which regions is the local correlation between these variables positive and where negative?
  2. Interpret the climatological meaning of these relationships.
3789 3 The attached archive contains long term mean precipitation values in autumn (September, October, November) and in spring (March, April, May) measured in the Baltic meteorological stations.
  1. Calculate spatial correlogram between these variables at 10 km wide distance intervals up to 500 km. Add the correlogram to the answer.
  2. At which interval is the spatial correlation between spring and autumn precipitation amount statistically significant at p < 0.01 level?
  3. How to interpret this correlation climatologically?