GEOB 270 Jan 2018

Lab 3: Planning for flooding/Storm Surge

Spatial Analysis, Tables, Editing

Due: 2 week lab, due at the beginning of your lab time in 2 weeks. Any labs submitted after the start of your lab time, will be marked late. Lab out of 38

Introduction

In this lab, you will be working more with spatial and tabular datasets, both raster and vector. You will be reviewing (from lab 2)spatial and raster data, coordinates systems, projections, and then be introduced to a variety of basic GIS Analysis tools.

You will be given a scenario for which you will perform GIS analyses. You will identify:

  1. the scenario or problem to be solved;
  2. the data you need;
  3. the geographic analysis to be performed;the analytical operation or GIS tools you need to access in ArcGIS
  4. the output maps of the results of the analysis
  5. data integrity (accuracy, etc)

This lab and lab 4 and lab 5 will follow a similar 5 step process which will also be the process that you will come up with for your final projects.

1. Scenario (problem to be solved)

You are a GIS analyst with Metro Vancouver (all municipalities in the Vancouver Region).

Given the geological and climate change weather events that have transpired recently around the world, including recently in Tofino on Vancouver Island with the waves and flooding, your supervisor wants you to create a series of maps that highlight the areas in the Metro Vancouver that are at risk to flooding or a storm surge.

What are Storm Surges? - a rising of the sea as a result of atmospheric pressure changes and wind associated with a storm.From Wikipedia - storm surges:

In the United States, one of the greatest recorded storm surges was generated byHurricane Katrinain 2005, which produced a maximum storm surge of more than 25ft (8m) in southernMississippi, with a storm surge height of 27.8ft (8.5m) inPass Christian.Another record storm surge occurred in this same area fromHurricane Camillein 1969, with a storm tide of 24.6ft (7.5m), also at Pass Christian. A storm surge of 14ft (4.2m) occurred inNew York CityduringHurricane Sandyin October 2012.

Although the height, speed and inland reach of a storm surge and resultant flooding depend on many factors (winds, tides, inland waterways such are the Fraser River and False Creek, etc) we will simplify for this exercise that the immediate danger zone from a storm surge is mainly to the low elevation areas at, or below, 10 metres that lie up to 1kilometre inland from the shoreline.

You are to develop 5 maps and 1 table for your supervisor:

1. and 2. overview map of map Metro Vancouver and of all areas of potential storm surge flooding – elevation from 0-10 meters, colour coded by elevation; second map done in layout view with legend and title.

3. map of potential flooding: refined analysis of potential storm surge by identifying areas of potential flooding by elevation within 1km of shoreline

4. map of all roads affected by potential flooding, colour coded by type, with a table on the map summarized road type by total length

5. False Creek tidal flats area detail map: finer resolution elevation information, map of schools, and new St. Paul’s hospital location that could be affected by potential flooding

2. The Data

Given the scenario and the expected deliverables (maps and tables above), in any GIS project, you should be creating a table of data that is necessary for this geographic analysis task.

Layer Name / Description / Raster/Vector / Feature class / Attributes / source
dem92g / DEM, 25m resolution / Raster tiff / Value (elevation) / Open data Canada
fcdem1m / DEM, 1m resolution, resamples from .5m / Raster tiff / Value (elevation) / City of Vancouver
shoreline / Outline of Metro Vancouver / vector / line / n/a / Statistics Canada
roads / Roads in Metro Vancouver / Vector / line / Road_type / CanMap
education / Schools in Metro Vancouver / Vector / point / Name, type / CanMap
van_schools
(download, not in Getdata) / Public schools in City of Vancouver / vector / point / City of Vancouver

Some of the data for this lab is open-data, in other words, free data available from the federal or municipal governments. There is one dataset we are using that is not open or public data, that UBC has a licence to use, and that is DMTI CanMap data.

Open Data Canada:

- course resolution DEM, clipped to MetroVancouver;

Statistic Canada/Census Canada:

- shoreline - geography file for Metro Vancouver Census Metropolotin Areas (CMA)

City of Vancouver:

- fine resolution DEM of City from Lidar data (data was modified for your lab: DEM was clipped to False Creek area because size of DEM very large for whole class to be working with and storing on computers)

- schools downloaded and renamed to van_schools

CanMAp data: in lab3 getdata – not open data

- Some of the data has been downloaded and bundled for you into a lab3 directory, some of the data you will access and download yourself.

Loading the Data

GetData: the file you load with GetData is zipped for compression as it contains 2 raster DEMs that are large files. You need to run extractall and set the path to C:\temp

IMPORTANT: when you need to leave the lab, zip all the data in lab3 in C:\temp and copy the .zip file to H:

The dataset downloaded contains a geodatabase of shapefiles and 2 raster DEM or tiff files. Geodatabases provide a file management structure for the geo-spatial data in ArcGIS. With whatever GIS software you are using, it is always important for project management to carefully manage and store your data. Geodatabases offer structural, performance and data management advantages over a collection of shapefiles in a directory. In this lab, we have given you the data in a geodatabase. In the next lab, you will create a geodatabase.

Launch ArcCatalog as a standalone program and preview the contents of C:\temp\lab3\

QUESTION 1(3)

What is the coordinate system, projection, linear units, and datum for your geodatabase?

What is the resolution of the dem92g (.tiff)?How accurate do you think this data is?

Close ArcCatalog.

Preparing a Geodatabase for Analysis

Launch ArcMap. In the Getting Started window select the Blank Map from My Templates and set your default geodatabase to C:\temp\lab3\analysis.gdb. Click on OK to begin. Save your map as c:\temp\lab3\mylab3.mxd.

Setting your default geodatabase to C:\temp\lab3\analysis.gdb will help you avoid losing your data. It will insure that the output of any analytical operation is stored in analysis.gdb as a feature class or a raster dataset, unless another location is chosen.

3. The Geographic Analysis: determine areas for possible storm surge flooding

a)Visually or cartographically classify raster DEM to show low elevation areas

b)Create potential storm surge flood map: Low elevation within 1km shoreline:

  1. Buffer shoreline by 1km: Perform buffer proximity analysis
  2. reclassify raster layer to get only low elevation
  3. convert raster low elevation to vector
  4. Combine the 1 km buffer and the lowlands (10m and under) with polygon overlay tool intersectto create area of potential flood from storm surge (low elevation, within 1km shore
  5. Create map of potential flood

c)Create map and summary table of roads that fall within potential flood

  1. create table of total length of highways and expressways within potential flood
  2. create map with major roads in potential flood highlighted, and summary table

d)False Creek Tidal Flats development: Detailed analysis of potential flood

  1. Zoom to detailed DEM of False Creek and create map of area symbolized
  2. Download schools from City of Vancouver open data site for display on map
  3. Digitize new layer of outline of new site for St. Paul’s Hospital
  4. Create map of detailed area

a) Visually or cartographically classify raster DEM (elevation) to display low elevation

We need to classify and symbolize the DEM to meaningful values and colours that represent low elevation flooding. Based upon historical data for storm surges, we will classify the elevation pixels (VALUE) into potential storm surge values that represent an elevation of 0-1 meter, 1-2 meters, 2-3 meters, 3-5 meters, 5-10meters to represent potential flooding, then 10-20 and anything over 20 as non-flooding risk areas. We will take the individual cell pixel values of elevation, and group them into these 8 categories for display purposes.

Bring up dem92g layer properties

Layer > Properties > Symbology

Change default symbology from continuous colour ramp to classified.

Under show, Click on Classified

Change default from 5 to 8 classes

Click on Classes

Change classes to 8

Don’t worry about the colour ramp

Click on Classify…

Under Break Values, change the default values to the following:

0

1

2

3

5

10

20

1,986

Click OK

Change colours individually of the classes.

Click on colour box beside 0

Change colour to light gray

OK

Change colours to shades of blue for 0-1, 1-2, 2-3, 3-5, 5-10

Change colour to a green for 10-20 and then yellow for 20-1,986

Click OK.

Zoom and pan to get metro Vancouver in the centre

Save your work as lab3_lowelevation.mxd

QUESTION 2 (3)

Export your map to lab3_map1lowelevation.pdf

submit

Create a map with a legend and a title:

Go to map or layout view

Under View, click on Layout View

(if you right-click on the non-map/layout white space around the map, under page and print set up you can change orientation from portrait to landscape if you wish).

Insert a title and legend on your map

Under Insert

Title – add a relevant title to your map

Legend: ensure you only have a legend for dem92g

Under title for legend put in Elevation Meters

Include all of the standard map elements (title, legend, scale bar, name + affiliation + date, north arrow, and data source, and coordinate system info).

Include all these map elements for any subsequent maps you create in this course.

QUESTION 3 (5)

Export map to a pdf, call it lab3_map2lowelevleg.pdf

b)Create potential storm surge flood map: Low elevation within 1km shoreline:

  1. Buffer shoreline by 1km: Perform buffer proximity analysis
  2. reclassify raster layer to get only low elevation
  3. convert raster low elevation to vector
  4. Combine the 1 km buffer and the lowlands (10m and under) with polygon overlay tool intersect to create area of potential flood from storm surge (low elevation, within 1km shore
  5. Create map of potential flood

Visually, we have created a map of potential flooding. If we want to further refine our analysis to look at data of a certain elevation, within 1km of the shoreline, we need to do some geographic analysis with the data.

  1. Buffer shoreline by 1km: Perform buffer proximity analysis

The first condition of our analysis isto find areas within 1kmof the shoreline. To do this, we will use the Buffer tool.

  • Add the shoreline feature class to your empty map.
  • Open the ArcToolbox window and expand the Analysis Tools by clicking on the ‘+’ sign next to the toolbox.
  • Expand the Proximity toolset and double-click on the Buffer tool.
  • To create a 1kmbuffer around the shoreline, enter the following parameters and press OK:
  • Input features: shoreline
  • Output feature class: C:\temp\lab3\analysis.gdb\shore1km
  • Linear unit: meters
  • Dissolve type: ALL

Every tool dialog box has a Help section that can be toggled on/off by using the Show Help > or < Hide Help buttons in the bottom-right of the tool’s dialog window. As you enter the different parameters, the help content changes to explain what each parameter means.

  • Click OK.

After a moment, the new shore1km layer is added to the Table of Contents (TOC).

NOTE

Notice that the buffer extends on both sides of the shoreline.

QUESTION 4 (1)

What does 1km represent; each side of the shoreline or the total width (use the measure tool)?

Notice as well that the buffer extends into the mountains, think about why this is, and whether it affect our analysis. This is an example of how you need to be aware of what is happening during GIS analysis and not just blindly click on buttons.

  1. reclassify raster layer to get only low elevation

The second condition in our analysis is that the land affected must be at or below 10 metres.

  • dem92g - open its attribute table. Notice the number of pixels that exist at or below 10m

From Question 1

What is the resolution of the DEM dem92g?

(Hint: remember that units are important when expressing resolution)

QUESTION 5(2)

How many pixels have values of 10 metres or less in the Metro VancouverDEM? How much area in km2 does that represent?

(Hint: use the Statistics command)

For our analysis, we don’t really need the variety of elevation values. What we would like to do is to renumber all the low values as ‘1’ and the higher values to ‘NoData’. To do this simplification, we will use the Reclassify tool:

  • Make sure the Spatial Analyst extension is checked on under Customize >

Extensions…

  • Bring up the ArcToolbox window and select Spatial Analyst Tools > Reclass >Reclassify
  • In the Reclassify dialog window, enter the following parameters and press OK:
  • Input raster: dem92g
  • Reclass field: VALUE
  • Click on the Classify button
  • Under Method:Equal Interval(this will allow us to change thenumber of categories)
  • Classes:2
  • Under Break Values: type10for the first value and click OK
  • You should now have only 3 rows in the reclassification table. For the second row, change the New Value from ‘2’ to ‘NoData’
  • Set the output raster to C:\temp\lab3\analysis.gdb\lowelevation
  • Click OK.

The lowlands raster layer is now added to the TOC. It contains values of ‘1’ for the areas under threat and nothing should be visible for the other areas.

  1. convert raster low elevation to vector

The areas we are searching for are those that are within a) the 1 km buffer and b) below 10 m in elevation. We need to combine these layers to isolate the areas common to both (i.e. the areas susceptible to the storm surge/flooding).

You can easily identify their intersection cartographically. To do this:

  • Turn off all layers except shore1km and lowelevation. Make sure the buffer is above the lowlands raster.
  • Open the properties for shore1km and under the Display tab enter a Transparency value of 30% and click OK.

The common areas between both layers should be easy to spot. However, in order to use this information in our analysis, we need to create a data layer containing these intersections. We first need to convert the lowelevation raster into vector:

  • In ArcToolbox, select Conversion ToolsFrom RasterRaster to polygon. Enter the following parameters and click OK. The new vector layer is added to the TOC:
  • Input raster: lowlands
  • Field: Value
  • Output polygon features: C:\temp\lab3\analysis.gdb\lowelevation_vector
  • Ensure that Simplify polygons is NOT checked

If you have not done so yet, this would be a good time to save your map document. Make sure to save it to C:\temp\lab3\ and NOT inside the analysis.gdb. The geodatabase should only contain actual data files.

  1. Combine the 1 km buffer and the lowelevation_vector (10m and under) with polygon overlay tool intersect to create area of potential flood from storm surge (low elevation, within 1km shore)

To create a vector layer that satisfies both conditions (area susceptible to a storm surge/flooding: within 1km buffer, and <10 m elevation) we will combine them usingthe overlay tool called Intersect:

  • In ArcToolbox, select Analysis toolsOverlayIntersect
  • Enter the following parameters and click OK:
  • Features: lowelevation_vector,shore1km
  • Output Feature Class: C:\temp\lab3\analysis.gdb\flooding
  • Accept all other default values
  • After the new layer has been added to the TOC, turn off the lowelevation_vector

shore1km layer

save your work to lab3_flooding.mxd

  1. Create map of potential flood

Go to map or layout view

Under View, click on Layout View

(if you right-click on the non-map/layout white space around the map, under page and print set up you can change orientation from portrait to landscape if you wish).

Insert a title and legend on your map

Under Insert

Title – add a relevant title to your map

Legend: ensure you only have a legend for dem92g and flooding

QUESTION 6 (4)

Export map to a pdf, call it lab3_map3flooding.pdf

c)Create map and summary table of roads that fall within potential flood

  1. create table of total length of highways and expressways within potential flood
  2. create map with major roads in potential flood highlighted, and summary table

One of the advantages of GIS is that you can select features in one layer based on their location relative to features in a different layer. Since we have a polygon of flooding, and a layer of roads, we can useSelect by Location to display all the roads that fall within the flooding areas.

  • Add Roads layer
  • In the ArcMap menu bar, click on SelectionSelect by Location
  • Input the following parameters before pressing OK
  • Selection method: select features from
  • Under Target layers put a check next to roads
  • Source layer: flooding
  • Spatial selection method for target layer feature(s): intersect the sourcelayer feature

You will see all the blue roads that intersect the flooding areas are highlighted as selected. We will create an independent layer for these selected roads

  • Right-click on roadsSelection create layer from selected features
  • a new layers have been added to the TOC, roads selectionyou should only see roads within the flood layer

Create a table of total length of roads in the potential flooded area, summarize by road type.

  • Open the attribute table for roads selection
  • View the attributes
  • Road_type indicated the code for highways etc At the end of this lab is a file which explains the Road_type code.
  • For the roads selection layer, symbolize expressways and highways (ROAD_TYPE 1 or 2) in a thick black or red line
  • Right click on roads selection
  • Symbology
  • Click on categories, pick ROAD_TYPE as value, click on add all values
  • Double click on 1 and 2 and change symbol to thicker black line or red line.
  • For roads 3 4 5 leave as fine thin black line
  • Create a summary table of the total length for each road type that is affected by potential flooding:
  • Right click on road type, summarize by shape_length, sum