540 likes | 723 Views
LULC & Fractional cover change in Inner Mongolia (1992 –2005). Ranjeet John LEES Lab Dept of EEES , U Toledo. Objectives.
E N D
LULC & Fractional cover change in Inner Mongolia (1992 –2005) Ranjeet John LEES Lab Dept of EEES, U Toledo
Objectives • To quantify the degradation of grasslands and other dominant land cover types in Inner Mongoliaasa function of fractional cover ( % canopy) change between 1992 and 2005
Study Area • Inner Mongolia lies between 37°01’ and 53°02’North Latitude and 95°02' and 123°37' East Longitude, third largest province in China (1.18 million km2 ) • bounded by Heilonjiang, Jilin, Liaoning, Hebei, Shanxi, Shaanxi, Ningxia Hui Autonomous Region, Gansu & Xinjiang • The principal mountain ranges are the Greater Hingaan in the east and the Yinshan and Langshan in the central part • Deserts include the Gobi desert in the north , Mu Us and Hobq deserts, south of the Huang He (river); the Tengger & Badain Jarian desert in the West
RATIONALE • Grasslands comprise most of land area (38.2%, 2005 MODIS IGBP ) and have increased in extent (33.26%, 1992-930) • fc is not only indicator of biomass but also imp. biophysical variable that controls hydrological processes in arid/semi arid regions (Shuttleworth, 1995, Goodrich et al, 1995) • Fractional cover (fc) has been derived and studied at multiple scales before (Li X.,2003) but change over time (and across LULC types) not documented
RATIONALE • Increase in croplands between 1992 (7.37%) and 2005 (10.80%); very controversial subject, as the general idea is that farmland is being abandoned though other studies show that it has expanded(Liu, 2005) • this change in land use is critical as it may modify the local climate (Bounoua, 2003) through replacement of native grasslands(include drought adapted C-4) and forests with crops, which transpire less and thus leads to warming
Goats in vegetation-starved areas must wear protective clothing to keep other desperately hungry goats from eating their hair. http://www.greenpeace.org/international/news/desertification-in-inner-mongo
Chinese documentarian Lu Tongjin has been chronicling desertification in Mongolia since 1995 http://www.greenpeace.org/international/photosvideos/slideshows/livelihoods-blown-away
Taklimakan desert Jan 1-8 2005 http://earthobservatory.nasa.gov/Newsroom/
Taklimakan desert Jan 1-8 2006 http://earthobservatory.nasa.gov/Newsroom/
Increase in xerophytous grasshopper species (Kang & Chen, 1995) Decrease in ET (Bounoua, 2003) Desertification Climate drier (Eastman et al, 2001) Reduction in Grassland fc Increase in crop area Livestock grazing Increase in population Decrease in precip Soil Erosion Aerosols
Research Question • What is the change in % canopy cover of grasslands between 1992 and 2005? • Can we measure the intra-seasonal dynamics of fractional cover for grasslands and other LULC types?
Rangeland Steppe Community • Typical Steppes & Meadow Steppes are sub types found in Inner Mongolia, most commonly used for grazing and animal produce • Typical Steppe developed under semi-arid conditions with annual precip under 350m, capable of drought tolerance, include Stipa, Leymus, Artemisia, & Festucca • Meadow Steppe developed in areas with moist fertile soils rich in organic matter with annual precip of 450mm, include Stipa baicalensis, Leymus chinenesis, Cleistogenes mucronata, etc (Sun, 2005)
DATA ACQUISITION AND PRE-PROCESSING • MOD13Q1:MODIS derived 16-day composite Vegetation Indices Global dataset 1km, used NDVI VI product to obtain green fractional cover • AVHRR 1km NDVI (1992-93) 10 day aggregates • MOD12Q1: MODIS derived LULC (IGBP) • AVHRR derived IGBP (1992-93) • MOD11A2: Land Surface Temperature (8 day, 1km) • SRTM 90m & misc shapefiles (extent, flux tower sites, etc)
Cross-tabulation of AVHRR IGBP 1992 (columns) against MODIS IGBP 2005(rows) 11permanent wetlands 12 croplands 13 urban and built-up 14 cropland/natural vegetation mosaic 1 evergreen needleleaf forest 2 evergreen broadleaf forest 3 deciduous needleleaf forest 4 deciduous broadleaf forest 5 mixed forests 6 closed shrublands 7 open shrublands 8 woody savannas 9 savannas 10 grasslands 15 snow 16 barren 17 water
Fractional Cover (Green and Senescent) Where SVI is the spectral vegetation index and SVIsoil andSVIfullcanopy are end-members - Qi, 2000
Annual average Fractional cover for all LULC types % fc reclassed into bins of 10% increment; 0-10 = 1, 11-20 = 2, etc
Cross-tabulation of fc1992 (columns) against fc 2005(rows) 1992-93 2005 % fc reclassed into bins of 10% increment; 0-10 = 1, 11-20 = 2, etc
Canopy fraction cover can be one of the indicators of degradation (in addition to landscape metrics) • Areas with sparse /no cover are represented by 0-2 • Undisturbed areas between 6 to 9 • Regrowth /succession between 2& 4 • There seems to be a shift towards less % cover when compared to % cover in 1992 • This has to be validated rigorously using another VI and scaling down to a higher resolution, landscape level (30m landsat scale as well as 250m EVI)
% fc reclassed into bins of 10% increment; 0-10 = 1, 11-20 = 2, etc
% fc reclassed into bins of 10% increment; 0-10 = 1, 11-20 = 2, etc
Monthly % Canopy cover dynamics for grassland cover 2005 1992
Monthly % Canopy cover dynamics for grassland cover 2005 1992
Monthly % Canopy cover dynamics for cropland cover 2005 1992
Variation in fc might be explained by • Variation in vegetation water content • Precip • Population (within county/province) • Distance to villages/towns
FUTURE WORK • Process and analyze vegetation water content using 3 indices; NDSVI, LSWI & NDWI • variability in water stress in context of cover type: protected lands vs disturbed lands (ie, low % fc) • Repeat fc work with EVI to get a better estimate • The water stress maps might explain variation in fc • Model AET using Nishida (2003) / Senay (2005) • Change in GPP & NPP (time/space) per cover type
Vegetation Indices • Senescent Vegetation Index: NDSVI = (ρswir – ρred) / (ρswir + ρred) - Qi et al., 2002 • Enhanced Vegetation Indices: ρNIR, ρRed and ρBlue = atmospherically corrected surface reflectance L = canopy background brightness correction factor (1) C1 and C2 = atmospheric resistance Red and Blue correction coefficients (6&7.5) G = Gain factor (2.5)- Huete et al., 2002
Acknowledgments Terra Orbit • Dr Jiquan Chen • Dr Asko Noormets • NASA NEWS 2004 NRA: NN-H-04-Z-YS-005-N • Rest of LEES Lab http://modis.gsfc.nasa.gov/gallery/#
References • Bounoua, L., R. S. Defries, M. L. Imhoff, and M. K. Steininger, Land use and local climate: A case study near Santa Cruz, Bolivia, Meteor. & Atmos. Physics, ISSN: 0177-7971 (paper) 1436-5065 (on line), 2003 • Eastman, Joseph L., Coughenour, Mike B., Pielke, Roger A.Does Grazing Affect Regional Climate? Journal of Hydrometeorology 2001 2: 243-253 • Kang, L. & Chen, Y. –L. 1995 Dynamics of grasshopper communites under different grazing intensities in Inner Mongolian steppes. Entomologia Sinica 2, 265–281 • Li, X. –B., Chen, Y. –H., Shi, P. –J. & Chen, J. 2003 Detecting vegetation fractional coverage of typical steppe in Northern China based on multi-scale remotely sensed data. Acta Botanica Sinica 45, 1146–1156. • Liu, J., Liu, M., Tian, H., Zhuang, D., Zhang, Z., Zhang, W., Tang, X., Deng, X., Spatial and temporal patterns of China’s cropland during 1990–2000: An analysis based on Landsat TM data, Remote Sensing of Environment 98 (2005) 442 – 456
List of Abbreviations/Definitions • Fractional cover: Percent canopy cover • MODIS: Moderate Resolution Imaging Spectroradiometer • AVHRR: Advanced Very High Resolution Radiometer • NDVI: Normalized Difference Vegetation Index • EVI: Enhanced Vegetation Index • NDSVI: Normalized Difference Vegetation Index • LSWI: Land Surface Water Index • NDWI: Normalized Difference Water Index • MOD09A1:Surface Reflectance 500m (VIS, NIR, SWIR) • MOD12Q1: MODIS land cover 1km • MOD13Q1: Vegetation Index 250m • MOD11A2: Land surface temperature 1km • MOD17A2: GPP/NPP 1km • IGBP: International Geosphere Biosphere Program