The purpose of this study was to evaluate the spatial distribution of geochemistry in Swedish till and whether interpolations could predict unknown values of geochemistry between samples at a national spatial level, Sweden, and a regional, the county of Västerbotten. The information of the spatial distribution of the elements and the interpolations accurateness has several applications. For example, the establishing of infrastructure. Which, in Sweden, is regulated by Naturvårdsverkets guidelines for sensitive land management (KM) and less sensitive land management (MKM) since elements can be harmful for the environment and health above certain levels. The guidelines for waste disposal of soil with levels of less than slight risk (MRR) does also acquire knowledge about the background levels of an area which are found in the C-horizon. The aim of the study was therefore to answer the following questions 1) Which interpolation method provides the most accurate prediction for the various elements at the national and regional level? 2) How does the calculated levels of geochemistry differ from the sampled ones for the two spatial levels? The study was conducted by studying the levels of ten elements in the C-horizon at the two spatial levels were the regional had greater sample density than the national. The interpolations that were used for these elements were the local interpolation methods kriging, inverse distance weighting (IDW), natural neighbour, thiessen polygons and triangular irregular network (TIN). Samples were gathered from Sweden’s geological Surveys (SGU) for the elements that are regulated by Naturvårdsverkets guidelines which was why these ten elements were studied: arsenic, barium, cobalt, chromium, copper, nickel, lead, antimony, vanadium, and zinc. The interpolations were done with two thirds of the data. A validation was done with the remaining third by calculating root mean square error (RMSE). For the interpolation method with the lowest RMSE, the mean absolute square error (MAPE) was calculated for all the validation points to see how the calculated levels differed from the samples. The result showed that kriging and IDW were the most accurate interpolation methods for the data but that some of the studied elements need even greater sample density to become more correct. This can be solved by doing a cross-validation of the existing data. Furthermore, the interpolations were more accurate at the regional level for elements except antimony. The higher accuracy can be explained by the higher sample density at the regional level. At the national level the interpolations worked better in the north of Sweden than in the south which needs further studying. Overall, the interpolations were the least accurate when the levels of the elements were low, which may be the reason why antimony showed higher RMSE at the regional level. In conclusion the study showed that it is possible to use interpolations to predict values at unknown places with different accurateness.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:umu-210308 |
| Date | January 2023 |
| Creators | Hansen Österlund, Sara Emilie |
| Publisher | Umeå universitet, Institutionen för ekologi, miljö och geovetenskap |
| Source Sets | DiVA Archive at Upsalla University |
| Language | Swedish |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0026 seconds