Remote monitoring of a coastal wetland in Lake Erie - PowerPoint PPT Presentation

1 / 33
About This Presentation
Title:

Remote monitoring of a coastal wetland in Lake Erie

Description:

Remote monitoring of a coastal wetland in Lake Erie – PowerPoint PPT presentation

Number of Views:25
Avg rating:3.0/5.0
Slides: 34
Provided by: Nisha2
Category:

less

Transcript and Presenter's Notes

Title: Remote monitoring of a coastal wetland in Lake Erie


1
Remote monitoring of a coastal wetland in Lake
Erie
Nishanthi Wijekoon Kent State University, OH
2
(No Transcript)
3
  • Overview
  • Using Landsat5 TM data (2005)
  • Generating a calibration curve
    to quantify total suspended particulate
    (TSP) concentration during summer
  • Varimax-rotated Principle component analysis
    to
    identify the distribution of
    ? phytoplankton
    ? Macrophytes
    ? Terrigenous component -
    suspended sediment
    - exposure of mud flats

4
Vegetation types in OWC estuary
Other terrestrial vegetation
Phragmites
  • Image from 6-30-2005

5
(No Transcript)
6
(No Transcript)
7
Dynamic barrier beach- summer 2005
8
Sample collection Laboratory Analysis (Summer
2005)
  • Samples collected at the same days when Landsat 5
    TM satellite over-passes the area (once every 16
    days)
  • GPS referenced15 sampling stations along the
    creek

9
  • Parameters measured
  • ?Total Suspended particulate (TSP)
    concentration
  • (Instrument - Malvern Masterizer 2000)
  • ? Chlorophyll-a absorption (Chl-a)
  • (Instrument - LabSpecPro FR
    Spectrophotometer)

10
Malvern Masterizer 2000
Total Suspended particulate (TSP) concentration
11
LabSpecPro FR Spectrophotometer
Chlorophyll-a absorption
12
Satellite Image Analysis
  • Software - PCI Geomatica 10.0
  • Image source - www.ohioview.org
  • Used back ground corrected Landsat 5 TM images
  • Remote sensing Indeces
  • calculated regression coefficients for NDWI
  • Normalized Difference Water Index
  • Normalized Surface Wetness Index
  • Principle Component Analysis (varimax rotated)

13
Landsat5 TM bands used for the study
Band number
14
Dark object subtracted 751 band combination
April-11 May-13 May-29
June-14 June-30
August-01 August-17
September-02 October-04

15
April-11 May-13 May-29
June-14 June-30
NSWI
August-01 August-17
September-02 October-04

16
(No Transcript)
17
(No Transcript)
18
Chlorophyll-a absorption (0.663 mm)
Absorption peak from
LabSpec Pro FR Spectrophotometer
Source Raven et al. 1976
19
  • Measurement of absorption peak- height at 0.663
    mm
  • Determine by linear interpolation

20
(No Transcript)
21
  • Principle Component Analysis (PCA) varimax
    rotated
  • ? Landsat5 TM images from April-11 to
    October-04 in 2005 (9 Images)
  • ? collected reflectance data from bands
    1-5 7 (6 bands)
  • Software - PCI Geomatica 10.0
  • ? PCA
  • Software - SPSS
  • ? Result Interpretation in spline
    technique
  • Software - ArcMap 9.0

22
Spectral pattern of rotated component matrix
Extraction method- Principle component
analysis Rotation method- Varimax with Kaiser
normalization
23
Factor-1 (Phytoplankton, variance 42.25)
April-11 May-13 May-29 June-14
June-30 Aug.-01 Aug.-17 Sep.-02
Oct.-04
24
Factor-2 (Terrigenous component, variance
31.41)
25
September-02
Factor 2 represents the exposure of mudflats
Star Island
26
October-04
Star Island
Star Island
27
Comparison of spectral patterns
Varimax-rotated Principle component analysis
LabSpec Pro spectrophotometer
28
Factor-3 (Macrophytes,variance 20.31)
29
Comparison of spectral patterns
Varimax-rotated Principle component analysis
LabSpec Pro spectrophotometer
30
Star Island
Spring
Mid-Summer
Fall
31
Further studies
  • Continue field studies in summer 2006
  • to validate the calibration data of NDWI and
    TSP concentration
  • to evaluate the factor identification of
    principal component analysis

32
Conclusion
  • A positive correlation exists between total
    suspended particulate (TSP) concentration and
    NDWI
  • Comparison of the results of principal component
    analysis (PCA) with spectral patterns and ground
    observations identifies the temporal and spatial
    distribution of
  • ? Factor-1 as phytoplankton (variance
    42) ?
    Factor-2 as terrigeneous component (variance 31)
    ? Factor-3 as macrophytes (variance 20)

33
Acknowledgements
  • Dr. Joseph Ortiz
  • Kent State University, OH
  • Funding NOAA NERR GR Fellowship
  • OWC Research Staff
  • (Frank Lopez, Dr. David Klarer)
Write a Comment
User Comments (0)
About PowerShow.com