What is it about?
An attempt is made to try to assess the error arising from using single point values to represent large areas. Correlation coefficients (r) greater than 0.86 were deemed necessary in order to sufficiently represent an area. Using several carefully selected stations across Texas, correlation values were computed using individual stations against their respective divisional averages, the statewide average, as well as between the single stations. Correlation values were found to be greatest during the winter months for both temperature and precipitation with the summer months exhibiting the lowest correlations. Also, station to station correlations were found to be directional dependent as inland stations exhibited low correlations when correlated with coastal stations. It appears that mean monthly values of temperature at a station can give a reasonable estimate of the larger area mean, especially during the cooler months and with minimal topography, while precipitation r values were so low as to discourage extrapolation from point to area.
Photo by Markus Spiske on Unsplash
Why is it important?
Climate change studies typically use point data of temperature and precipitation to represent larger areas. This study examines how much error can be expected from using just a few stations to represent a larger area. Essentially, it examines how reliable climate change models are in predicting the future.
Read the Original
This page is a summary of: A problem in climate change studies, extrapolation from point to area data: a case study in Texas, USA, International Journal of Environmental Studies, May 1996, Taylor & Francis, DOI: 10.1080/00207239608711039.
You can read the full text:
The following have contributed to this page