Working Title
Analysing of Land Cover and Land Use System Change in Mongolia with MODIS time series
Objectives
Assessment of land cover and land use system change based on remote sensing and comparison of change between and within different land use systems in Mongolia.
Contributions:
· Contribution to mapping land use systems based on remote sensing on a regional and national scale using MODIS.
· Contribution to the general topic of coping with desertification in Mongolia with the key questions: „where takes land degradation place“, („and at what intensity“).
· Contribution to the processing and integration of MODIS products in context of national desertification assessment.
Objectives:
1. Quality analysis of different global land cover type products
2. Generating reliable and robust time series of MODIS vegetation indices.
3. Development of additional criteria’s to differentiate different land cover type.
4. Developing of a model to map land cover type more accurate and different land use systems on a regional and national scale.
5. Analysing land use type and land use system change within and between categories in the last 10 years.
Methods
Method to reach objective 1:
Quality analyses of different global land cover type products
· Comparison of different Classifications, Availability and Quality of following global land cover types: Modis IGBP, Modis UMT, Meris Globecover.
· Validation of land cover type products with collected ground truth data.
Method to reach objective 2:
Generating reliable and robust time series of MODIS vegetation indices
· Downloading, mosaicing, projecting und clipping
· Nois-reduction with maximum value composite (MVC) and integration of a fitting function (Linear, Savitsky-Golay, …) with TIMESAT
· Weighting according to the quality of the pixel with TIMESAT
Method to reach objective 3:
Development of additional criteria’s to differentiate different land cover type (different types of grassland)
· Improvement and Adjustment of land cover types for the specific conditions in Mongolia.
· Analysing ground-truth data and the temporal change of the value of the vegetation indices (NDVI, EVI)
· Assumption: Different kinds of grassland and especially grassland with different quality have a different temporal development of Vegetation Indices during the year.
Method to reach objective 4:
Development of a model to map land cover type more accurate and different land use systems on a regional and national scale.
· Land cover type à Land use system
· Integration of more and other important factors according to the WOCAT geomatic-based model:
Factors: | Source |
Slope, Aspect, Hight | SRTM, ASTER GDEM |
Soil Typ | Natural Ressource Management / RS |
Temperature, Precipitation | Worldclim.org (1km2), VASClimO (0.5° lat/long) |
Population | ? |
Snow cover | MODIS Snow Cover MOD10A2 |
Life stock | Tropical Institut of Basel |
… |
· Mapping of specific land use systems with more precise land cover types and the development of NDVI- and EVI-values among the year.
Method to reach objective 5:
Analysing land use type and land use system change within and between categories in the last 10 years.
· LTA: Linear trend analysis
- General statistics: linear regression
- Escadafal, R. (frühstens 2005): Arid land cover change trend analysis with series of satellite images for desertification monitoring in Northern Africa.
- Baolin Li, Wanli Yu, JuanWang:An Analysis of Vegetation Change Trends and Their Causes in InnerMongolia, China from1982 to 2006
rK < -1 | Decreasing of vegetation cover | |
-1 > rK <1 | Stable vegetation cover | |
1 < rK | Increasing vegetation cover |
· STA: Seasonal Trend Analysis
- Harmonic regression
- Ronald Eastman, J. , Sangermano, Florencia , Ghimire, Bardan , Zhu, Honglei , Chen, Hao , Neeti, Neeti, Cai, Yongming , Machado, Elia A. and Crema, Stefano C.(2009): 'Seasonal trend analysis of image time series', International Journal of Remote Sensing
· BFAST: Breaks For Additive Seasonal and Trend
- Verbesselt, J. (2010): Detecting trend and seasonal changes in satellite image time series
Source: Verbesselet (2010)
