Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.sevin.ru/ecosys_services/references/Puzachenko_Forming%20of%20landscape%20structure%20on%20south%20of%20Valday%20highland.pdf Дата изменения: Thu Jun 10 18:28:18 2010 Дата индексирования: Sun Sep 12 11:32:05 2010 Кодировка:

Forming of landscape structure on south of Valday Hill
M.Yu. Puzachenko1, D.N. Kozlov2, E.V.Siunova
1 1

2

A.N. Severtsov Institute of Ecology and Evolution RAS, 119071, Moscow, Leninsky prospect 33, Russia, puzak@orc.ru Lomonosov Moscow State University Faculty of Geography, 119899, Moscow, Leninskiye Gory 1, Russia

South of Valday Hill (within Tver region borders, coordinates 55є - 57є N 30є-34є E) is a region of complicated spatial structure of landscapes which is determined by the border of Wurm glaciation maximum stage, relatively high amplitude of relief, its forms diversity (formed by geological base), deposits of Saalian and Wurm glaciations, continental gradient. Contemporary landscape structure is determined by high level of forest operational damage, historical conditions of land use which are changing during 20th century and nowadays due to withdrawal of fragmented areas from land use. We demonstrate methods and results of general landscape forming factors (parameters of order) determination, their physical sense and the scale of their influence to each component of landscape. In fact we assess the scale of mutual dependence of landscape components, dependent and independent variables of their condition. Landscape components reflected through spatial variation of numerous properties, determined on digital elevation model (made with topographic map 1:100000), composition of multispectral images Landsat 5, Landsat 7, Geocover (1990 and 2000) and gridded maps of climatic variables with 10 arc minute resolution. We used monthly average long-term variables for each month ­ precipitation, number of rainy days, average temperature, partial pressure of vapor, relative humidity, duration of sun shining, number of frosty days, wind speed. Detailed descriptions of landscape components made in 2600 points. Points lay on regular and selective schemes and GPS-positioned. Using spectral analysis of DEM the hierarchical levels of relief organization was revealed. Relief in each point characterized by relative elevation, slope and curvature for each level. Every property of every landscape component is presented via discrete units. Using discriminant analysis for each property spectral bands, vegetation indices and topographical variables transformed into independent factors. Regression calculated between factors and property. The property variation in space for each pixel can be described in 40-50% (2070%) cases. As a result we can create maps of this property and its factors. This method allows to explore each component and whole system. General dimension of all properties space do not exceed eight. Four of landscape forming factors describe 60% of variation of all properties. Each landscape component has properties independent from the whole and determining its autonomous variation. Set of the discovered regularity can be expressed by the following items: 1. Spatial structure of landscape is almost strongly fractal. Quasiperiodical component describes less than 5% of variation. Fractality is determined by mechanisms of geologic basis fractioning into blocks under geodynamic forces acceleration. 2. Quaternary sediments and their relief is also fractal and significantly determined by prequaternary basis. 3. Strong non-linear interactions can appear between mesoclimate, vegetation, soil, hydrosphere, agriculture and can generate mesoscale fractal structures. This inc reases dimension of landscape and increases quasiregular input into its structure. 4. Relations between components are determined well on macro level. The determination decreases to lower levels due to non-linear interactions. This means that we can not image more than three (reliably - two) hierarchical levels on one scale. The research is made with support of RFBR project # 03-05-64280.