350 likes | 462 Views
Assessing the Response of Coastal Marshes to Sea Level Rise at a Coast-Wide Scale. Michael S. Kearney Department of Geography and the Earth System Science Interdisciplinary Center University of Maryland College Park , MD 20742.
E N D
Assessing the Response of Coastal Marshes to Sea Level Rise at a Coast-Wide Scale Michael S. Kearney Department of Geography and the Earth System Science Interdisciplinary Center University of Maryland College Park , MD 20742 TM Marsh Index: Healthy, Mod Deterioration, Severe Deterioration
Canals in Holland regularly froze during ‘Little Ice Age’ (ca. 1400-1850) Winter Landscape, Pieter Breugel the Younger – 1601 Kunsthistorisches Museum, Vienna.
Chesapeake Bay has had a slow rateof sea level rise (~0.56 mm/yr) overthe last thousand years Kearney,M. 1996 J. Coastal Research 12: 977-983
Mean sea level has risen 30 cm(1 foot) over the last century PSMSL data Baltimore tide gauge
Processes of Marsh Loss • Local human activities: ditching, diking, berms, tidal restrictions (roads), Nutria over-grazing, groundwater withdrawals, subsidence, & etc. • Sea Level Rise: • -slow vertical accretion (low tidal flushing) • -plants dieback (especially during droughts) • - formation of rotten spots & then ponds • -ponds coalescence to open embayments
Physical changes in marsh condition accompany marsh loss Photo: J. C. Stevenson
Typical Spectral Curves of Vegetation, Soil, and Water Vegetation Soil Water
Simulated Thematic Mapper data from Scirpus americanus inundation experiment Pixel Value I Inundation Depth (cm)
Spectral Indices 1) Normalized Difference Water Index (NDWI): (Band 3 - Band 5) / (Band 3 + Band 5); 2) Normalized Difference Vegetation Index (NDVI): (Band 4 - Band 3) / (Band 4 + Band 3); 3) Normalized Difference Soil Index (NDSI): (Band 5 - Band 4) / (Band 5 + Band 4).
Image Preprocessing Image Indices Endmember Selection And Evaluation Raw Data Calibrated Data Band Y Band X NDSI NDVI NDWI Reflectance Conversion 1m CIR Photos Optimized Endmembers Calibrated Data NDXI Layered Indices Reflectance Wavelength Spectral Mixture Analysis And Thematic Classification NDXI Layered Indices SMA Fractions Multitemporal SMA RMSE Images Thematic Classification Data Processing Flow
Change In Marsh Surface Condition 1984-1993 For Delaware Bay
Change Detection at Larger Scales 1984 1993 1984 1993 Changes in marsh condition between 1984-1993 at Bombay Hook, Delaware based on the Landsat TM MSCI model results. Green = healthy to slightly degraded marsh; yellow = moderately degraded marsh; red = severely degraded marsh; blue = water. The area covered by the images is approximately 64 km2.
1993 1988 1999 2001 Water Upland Intact Moderate Severe Thematic class change: 1988 - 2001
Area Non-degraded Slightly to moderately degraded Severely to completely degraded Chesapeake Bay Upper and middle Bay* 31 (25,201) 50 (40,647) 19 (15,446) Lower Bay 28 (9,404) 52 (17,464) 20 (6,717) Delaware Bay North Shore (New Jersey) 38 (27,095) 43 (30,660) 19 (13,547) South Shore (Delaware) 55 (19,974) 35 (12,711) 10 (3,632) Marsh condition class percentages (hectares) for estuarine marshes in Chesapeake and Delaware Bays based on 1993 Thematic Mapper imagery.
CLIMATE CHANGE VULNERABILITY AND ADAPTATION The large scale features of climate change are well understood, but the projections of change at regional and smaller scales remain uncertain. • OPEN WATER AREAS • More extreme flows and stratification • Habitat loss • SHALLOW WATER AREAS • Habitat & SAV loss • Carbon sources • Total suspended solids • COASTAL WETLANDS • Loss of habitat and biodiversity • Functional loss • Loss coastal uplands
SE 590 Spectroradiometer (252 detectors over a range of 370-1100 nm) deployed 1.5 meters over marsh with video camera to record water depths in aluminum box
Landsat TM Mixture Model • Subpixel LOOK, Decomposes Each Pixel Into Vegetation, Soil, and Water Elements • w1fw + v1fv + s1fs = R1 • w2fw + v2 fv + s2fs = R2 • w3fw + v3fv + s3fs = R3 • The Indices are Independent, and Do Not Rely On Endmember Selection • -A CriticalIn General PCA-Based Mixture Models