Wykorzystanie zdjęć LANDSAT w badaniu struktury sieci osadniczej w wybranych krajach pozaeuropejskich

The main aim of the article is to assess the usage of the following methods for the settlement structure analysis: satellite images interpretation, Zipf rule and nearest neighbour index. This assessment is carried out during the analysis of the relationship between socio-economic development level and the diversification and complexity of settlement Settlement network development was especially investigated in the 1950. and 1960. R. B. Potter (1999) summarising heretofore research stated that the balanced and hierarchised settlement network does not develop in LEDCs. Therefore it was interesting to check if such a statement is not exaggerated. The analytical data were obtained from LANDSAT ETM+ images interpretation. Due to it, the coefficients of variation for particular countries were calculated to investigate settlement network diversification. The diversification was presented also on choropleth maps. Afterwards it was analysed if the Zipf rules (classic, using population number data and modified, using built-up areas values obtained from satellite images) describe the settlement network complexity in countries under investigation. The nearest neighbour index was used to check spatial complexity of the networks. By the networks diversification analysis, it was proved that there was a relationship between socio-economic level and networks diversity in the countries under investigation, although there are also other processes influencing this relationship. There is intense spatial dynamism in core regions and neighbouring peripheries in the countries with the most diverse networks (the most developed counties). In the least diverse countries there is the low spatial dynamism or there were significant decentralisation forces in the past. By the networks complexity analysis, the development level and networks complexity relationship was revealed. In the relatively less developed countries the settlement networks were far from equilibrium (primate city pattern) and built-up areas dispersed. In more developed countries settlement networks were close to equilibrium and concentrated, so compact urbanised areas were present. Changes in networks complexity take place in the most developed areas, however, peripheries are spread across the overwhelming part of the countries, which is irrespective to the level of their development. Hence, networks’ structures enhancement and their dispersion occur only in some parts of the countries. Therefore R. B. Potter’s and others statements seem to underestimate the LEDCs’ processes. The visual satellite images interpretation allowed the analysis of the data rarely used in settlement structure researches. This enabled the analysis with lack of population statistical data limitations, e.g. unreliable data not covering the whole population within the city or town, data not comparable between different countries (problem of different city definitions), from different years for each country. The comparison of modified and classic Zipf rules showed that the modification was correct. The data availability for modified rule was also much greater. This method occurred useful in LEDCs as spatial urbanised areas expansion is characteristic for the urbanisation process there (Cohen, 2006). The nearest neighbour index NNI analysis was also much more precise and due to the greater data availability, statistically significant. The conclusions of the methodological aim can be extrapolated, providing for the limitations described, for other countries not being covered by the research.