Methodology for creating historical land use databases : a case study for ICOS-station Hyltemossa, Sweden

Английский. This thesis studies how modern spatio temporal databases of land use changes can be created based on historical maps. The entire study was conducted based on the available maps over the study area, in southern Sweden. The research was conducted in such a manner that it presents to what extent can land use changes affect biodiversity and climate change. The thesis is not focused on what are the drivers for land use conversions, but rather how could changes in land use be easier identified when working with a set of (historical) maps over the same area. Based on a very first decision, in terms of what land use data can be used in further studies by different research communities, a study was conducted as to determinate what are the available historical sources. A database was created, containing land use data, classified accordingly to modern international land cover classification, extracted from historical maps, during the time frame 1812 - 2006. Considering the needs in terms of land use conversion for the biodiversity and carbon modeling communities, analyzing tools were developed to provide a fast and efficient way to easily identify in which time frame drastic land use conversions occurred and how much the study area was affected. The tools were created as to ease the preliminary analysis on land use maps, to quickly identify changes based on statistics, rather than visual analysis. The historical land-use database was tested over a study area in southern Sweden. The tool developed is able to extract statistics from at least 2 maps simultaneous and express how much of land class there is on each map. The users can quickly retrieve statistics when the surfaces for each land class are available.

Английский. According to the Intergovernmental Panel on Climate Change (IPCC, 2017), both land use conversions and use of fossil fuels are the sources of the increase of GHG, leading to global warming. Due to the high demand of land use conversions (such as increased areas of cultivated soil, forest clearance, urban areas development etc.), the land use suffers changes, some of them irreversible, whose effects can be measured in the amounts of greenhouse gas emissions released in the atmosphere. Furthermore, the land use and land cover change highly affect the local biodiversity, in terms of biodiversity loss or biodiversity alterations. Many of these changes were captured either in historical maps, in cadastral dossiers or written texts, by the authorized representatives in charge at that time. The GIS communities aim to process historical datasets in order to provide modern databases to other research communities, such the present study case. The methodology was structured into five processing steps. The first step was to identify the need for historical land use data, analyzed both from a carbon modeling and biodiversity perspective. The second part focused on checking the availability of historical land-use map resources, for the study area (Hyltemossa, Sweden), between the time frame 1812-2006. Once the needs and available map resources for such a database were known, the next step was to decide what land-use/land cover classification to use. It was chosen the modern Corine Land Cover Classification (Corine CLC). An important part was the actual creation of the historical land use database and the tools for analyzing the database, processing step number four. The historical maps were georeferenced and digitized, and stored as raster images into a database, using the GIS software ArcMap. To interpret changes over time, an analyzing tool was developed, with the capacity to detect changes from several overlaid maps of the same area (the fifth processing step). The function was developed from detecting pixel level changes to larger areas and the results are presented a table with statistics of a number of pixels that changed their values from one map to the other. The results and discussion focused on the resources used in this project, the digitized historical maps and the feasibility of the analyzing tool.