The Reduction of Blue Green Algae in Lake
The Reduction of Blue Green Algae in Lake Nosbonsing Land Classification on the Kaibuskong River Subwatershed 6/1/2013 White Winter. gis 1
Outline � � Project Team Study Area Project Statement Overview Methodology � Data � Mosaic � Unsupervised Classification � Statistical Analysis � � � 6/1/2013 Cost Analysis Challenges Recommendations White Winter. gis 2
Client �Scott Higgins � GIS Specialist � Department of Drinking Water Source Protection PROJECT TEAM Bennett White Janet Finlay Project Manager Project Advisor Hons. BA Wilfird Laurier B. Sc Mc. Master University Instructor at Niagara College Canada White Winter. gis 6/1/2013
6/1/2013 White Winter. gis 4
6/1/2013 White Winter. gis 5
Problem Statement � Recent Blue Green Algae Events � Lack of Information pertaining to land classes � Lake Nosbonsing and Kaibuskong Subwatershed PROJECT OBJECTIVE Identification of Land Classes Comparable Analysis Compare NEW LAND data against OLD LAND CLASS 6/1/2013 White Winter. gis 6
Benefits � Identification of land classes that have increased the abundance of phytoplankton in the watershed �Nutrient rich vegetation � Tangible items; �Updated land classes of unclassified subwatershed �Spatial Overlay Analysis �Hard and soft copy maps �Documentation 6/1/2013 White Winter. gis 7
METHODOLOGY v Data v Mosaic v Unsupervised v Spatial 6/1/2013 Overlay Analysis White Winter. gis 8
Data Quick. Bird Imagery Multispectral Imagery � 2 cm resolution � 271 Tiles � Forest Resource Inventory Data Panchromatic Imagery � Shape File � � Forest Type / Area � 1 tile = 130 mb � Mgmt. Practices 6/1/2013 White Winter. gis 9
Mosaic Process � 271 tiles � 130 mb per file � Final Image � 70 GB 6/1/2013 White Winter. gis 10
UNSUPERVISED CLASSIFICATION Unsupervised Classification ISODATA PCA 6/1/2013 White Winter. gis 11
Unsupervised Classification 11 Classes were selected � Algorithm splits and merges clusters � 6/1/2013 � ISODATA Algorithum �Iterative �Select �Organizing �Data �Analysis �Technique White Winter. gis 12
6/1/2013 White Winter. gis 13
Principal Component Analysis �Creates a new set of uncorrelated variables ○ Principle Components or eigenchannel �Linear Combinations of the original bands Makes the image smooth and clean, without data lose �Data reduction technique ○ Compresses a number of bands into a few components ○ Reduction in dimensionality of the data without a loss of information 6/1/2013 White Winter. gis 14
6/1/2013 White Winter. gis 15
UNSUPERVISED VS. Analysis PCAof 11 Classes 6/1/2013 White Winter. gis 16
6/1/2013 White Winter. gis 17
6/1/2013 White Winter. gis 18
SUPERVISED Training Sites (AOIs) Spectral Signatures 6/1/2013 White Winter. gis 19
Supervised Classification 6/1/2013 White Winter. gis 20
6/1/2013 White Winter. gis 21
MAXIMUM LIKELIHOOD • Input bands of data have a normal distribution -- Parametric Rule • Assumes that a pixel belongs to a particular class 6/1/2013 White Winter. gis 22
6/1/2013 White Winter. gis 23
OVERLAY ANALYSIS Old vs New Land Class Data 6/1/2013 White Winter. gis 24
6/1/2013 White Winter. gis 25
6/1/2013 White Winter. gis 26
6/1/2013 White Winter. gis 27
6/1/2013 White Winter. gis 28
PROJECT MANAGEMENT 6/1/2013 White Winter. gis 29
EVM (October 12 – June 14) Planned Value � End First Term Actual Value � � 8, 060. 00 � End Second Term � 8, 306. 01 � � 25, 494. 74 � June 14 End Second Term � 14, 742. 35 � � 48, 174. 00 6/1/2013 End First Term June 14 � 33, 375. 35 White Winter. gis 30
Cost Analysis � Baseline 1: December 14, 2012 Baseline 1: � Data Acquisition, Proposal � Baseline 2: � � $55, 393. 05 � Baseline 2: March 22, 2013 � $45, 409. 14 � Mosaic, Progress Report � Final Cost: June 10 th, 2013 � � $ 38, 038. 84 � Baseline 3: � Unsupervised & Supervised Cost Savings of $17, 354. 21 White Winter. gis classification, Statistical Analysis 6/1/2013 31
30 PM Management 11 23 Proposal Preparation 24. 5 30 17 49 Data Manipualtion Stage 1: Quickbird Progress Preparation Classifcation Part 1: Quickbird Data Manipulation Pat 2 FRI Classifcation part 2 FRI Statistics & Map Analysis 74 6/1/2013 Final preparation White Winter. gis 32
Challenges � Data & Storage Space � 1 tile= 130 mb, Mosaic= 70 GB � Processing Speed �Mosaic �Raster to Polygon Tool in Arc. GIS 6/1/2013 White Winter. gis 33
6/1/2013 White Winter. gis 34
Recommendations � Added Z-Values � Obtain appropriate method to transfer data �Niagara College Corporate Drop. Box Account � Determine 6/1/2013 the compatibilities between = White Winter. gis 35
Acknowledgments � North Bay Mattawa Conservation Authority � Ministry of Natural Resources (MNR) � Niagara College Canada �Instructors �IT Specialist (Colin Bissell) 6/1/2013 White Winter. gis 36
6/1/2013 White Winter. gis 37
Thank you �Questions � Bennett White �bennett. white 88@gmail. com �http: //whitewintergis. weebly. com 6/1/2013 White Winter. gis 38
- Slides: 38