Geospatial Intelligence - 2nd Test Notes PDF

# Geospatial Intelligence ## Teste - GIS lets you map where things move or the changing conditions in a place over time. Knowing what's changed can help you understand how things behave over time, anticipate future conditions, or evaluate the results of an action or policy. ## Topics - Why map ch...

# Geospatial Intelligence ## Teste - GIS lets you map where things move or the changing conditions in a place over time. Knowing what's changed can help you understand how things behave over time, anticipate future conditions, or evaluate the results of an action or policy. ## Topics - Why map change? - Defining your analysis. - Three ways of mapping change. - Creating a time series. - Creating a tracking map. - Measuring and mapping change. ## Why Map Change - To gain insight into how things behave. - To anticipate feature conditions/needs. - To decide on a course of action. - To evaluate the results of an action. (By mapping conditions before and after an action, you can see the impact.) ## Defining Your Analysis - You can map change by showing the location of features at each date. - You can calculate and map the difference in value for each feature between 2 or more dates. ## What You Need to Know Before Starting Your Analysis - Type of change you're dealing with. - Type of features you're dealing with. - How you're measuring time. - Type of information you need from the analysis. ## Types of Change | Data | Change | |---|---| | Change in location | Helps you to see how features behave so you can predict where they'll move. | | Change in character or magnitude | Shows you how conditions in a given place have changed. The change can be in the type of feature in a place or a quantity associated with each feature. | **Change in location and character are not mutually exclusive** ## **The Geographic Features** - The type of features you're mapping help you to determine the best method to map change. ### **Features That Move** #### **Discrete Features** - Can be tracked as they move through space. - They might be individual features, linear features, or areas (that usually expand and contract). #### **Events** - Geographic phenomena that occur at different locations. - The set of events can be tracked and mapped to show the movement of phenomena in a period of time. ## **Features That Change In Character or Magnitude** - Data summarized by totals, percentages or other quantities associated with features within defined areas. - Show the type of features in a place. - Can be represented by boundaries or as a surface. - Continuous quantities. - At any location, there is a measure of the values. The data is interpolated to create a surface. ## Measuring Time - The time pattern - A trend: Change between 2 or more dates or times. They indicate whether something is increasing or decreasing or the direction of a feature's movement. - Cycles: Show recurring patterns that reveal information about the behavior of the features. - Before and after conditions allow you to see the impact of an event. - Partitioning time - You can display feature locations or characteristics at 2 or more time periods. - Or you can summarize feature attributes over a period of time or several time periods. You'll also need to decide how many time periods to map and the interval between them. ## Comparing Methods | Method | Type of Change | Time Pattern | Pros | Cons | |---|---|---|---|---| | Time Series | Movement of change in character | Trend, Cycle, Before and after | Strong visual impact if change is substantial. Shows conditions at each date/time | Readers have to visually compare maps to see where and how much change occurred. | | Tracking Map | Movement | Trend, Cycle, Before and after | Easier to see movement and change with time series, especially if change is subtle | Can be difficult to read if more than a few features. | | Measuring Change in Character | Change in character | Trend, Before and after | Shows actual difference in amounts of values | Doesn't show actual conditions at each time. Change is calculated between 2 times only. | ## What Method to Use? - **Time Series:** If you want to show snapshots for 2 or more times, either movement or change in character. - **Tracking Map:** If you want to show feature movements between 2 or more recurring periods. - **Measure Change:** If you want to show the calculated difference in an attribute of a place between two time periods. ### Time Series | Change in | | Change in | |---|---|---| | Location | Tracking many individual features, like calls to 911, over time, or show movement like area boundaries. Spreading over time (ex: boundary of a wildfire). | Magnitude | Difficult to see the patterns of change between the maps if classification methods are different in each map. Quantile & equal interval are useful for comparing values over time. | Character | Categories can change from map to map so they may not be the same. Using the existing categories is more accurate but makes it harder to compare maps. | ## Number of Maps to Show - Showing fewer maps farther apart in time may make the change in values easier to see. - Showing more maps closer together in time may reveal patterns that are missed when using fewer maps. - However, It's difficult to compare more than 5 or 6 maps. - You can create maps and charts along with your maps to help show change. ## Tracking Map - **Mapping Individual Features** - If you want to emphasize the path the feature followed, draw a line to connect each date or time. The shorter the interval between features, the faster the movement. You can also show the change in character or magnitude of the features using different colors or symbols. - **Mapping Linear Features** - Often mapped before and after an event (ex: shoreline before and after a storm). - You can use colors or symbols to show change in character or magnitude of the feature. - **Mapping Contiguous Features** - To map the movement of a contiguous feature (area), draw the boundaries of the area at each time, or shade the areas using different colors or patterns. - **Mapping Events** - To show a trend in the movement of a phenomenon represented by discrete events, draw the events using a different color for each time period. - To show movement before and after an action or event, draw the events occurring before the target date in one color and the events after in another color. - To show movement over a cycle, code the events based on the period they fall in. ## Measuring Change - **Data summarized by area** - To show change in value for each area from a time period to another, we can create bar charts that show the area value in each time period, and insert them in the respective areas of the map. - To compare more than 2 time periods, we can create a trend line graph, one for each area, or one for multiple areas. ## Negative Values - After calculating the change values, there can be negative values, which means that the attribute has decreased in value. Generally, the negative values are represented in blue and the positives in orange and reds. ## What if the Boundaries Changed? - If you have historical data, the boundaries of the areas you're mapping may have changed. Small changes may not change the patterns on the map. In this case, we can ignore the changes or draw them in a less legible way on the map (ex: dashed line). - If the changes are big, it can be necessary to map the attributes in another way. A good solution is creating a continuous surface. ## Calculating Change in Continuous Categories - To calculate this change, it is necessary to use z different layers on the map, each of them with data from each time point to be analyzed. ## To Calculate the Change, We Can: - Create a map that only shows the areas that were changed. - Calculate the area change for each category and present the values in a chart. - Calculate the quantity of change from one category to another and present a chart. ## In the First Approach, We Overly the 2 Layers to Create a New Layer That is the Junction of Those 2. Then, We Select Only the Parts Where the 2 Layers Are Different. ## In the Second Approach, We Sum, the Area for Each Category for Each Date. Calculate the Change by Subtracting the Area for the Original Date From The Area at the Second Date and Assigning it to a New Column. This Method Doesn't Create a New Layer. You'll Need to Display the Information Using a Table or Chart, Along With the Maps From Both Dates. ## In the Last Approach, You Can Use Either Vector or Raster Data to Calculate How Much of Each Original Category Has Changed to Each New Category. ## Calculating Change In Continuous Values - To create a change map between two surfaces, you subtract the layers. The GIS calculates the difference between each cell on the first layer and the corresponding cell on the second layer. ## GIS-Based Multicriteria Evaluation (MCE) ### Introduction - Selecting the best places for a new business/infrastructure is a complex socio-technical process. - Decision-making process involves clarifying and resolving problems through information exchange and negotiation among stakeholders. - GIS and spatial analysis are not enough. - Need to integrate values of decision-makers and policy and management goals to evaluate alternatives and impacts. - MCE provides a framework to facilitate this approach. ### **Basic Principles of GIS-Based MCE** #### Key Components - **Decision Alternatives** - **Criteria:** Guidelines or requirements used as a basis for decision. - **Factors:** (ex: distance to road - near = most suitable, far = less suitable). - **Constraints:** (ex: protected area, water body, etc.). - **Decision Rule:** Numerical method for prioritizing alternatives according to how well they satisfy the criteria. #### Example **Goal: Riparian Vegetation Management** - **Nature Conservation Value** - **Profit From Changing Land Use** - **Potential Flood Damage** - **Soil Erosion** - **Effects on Landscape** **Alternatives** - No Clearing. - Selective Clearing. - Total Clearing. ### Performance Matrix | (Weights) | Conservation (0.21) | Profit (0.10) | Flood (0.12) | Erosion (0.18) | Landscape (0.39) | |---|---|---|---|---|---| | No clearing | 1 | 0 | 0.8 | 0.9 | 0.3 | | Selective clearing | 0.7 | 0.5 | 0.6 | 0.4 | 0.9 | | Total clearing | 0.1 | 0.9 | 0.2 | 0.1 | 0.2 | ## Performance Value - Performance values are determined based on facts, scientific evidence, stakeholders’ judgments. In this case, the scale is from 0 to 1, but can also be from 0 to any other value. - **Weights** reflect stakeholders’ judgments. The sum of the weights should be equal to 1. - **Weighted summation:** $$Sj = \sum_{i=1}^n W_i X_{ij} $$ - **Ranking the Alternatives** - No Clearing: 0.21x1 + 0.10x0 + 0.12x0.8 + 0.18x0.9 + 0.39x0.3 = 0.58 - Selective Clearing: 0.69 - Total Clearing: 0.23 - Highest value, or the one with the highest performance value. ## Methods of MCE - **6-Step Procedure** 1. **Establishing the Decision context.** 2. **Structuring the decision problem.** 3. **Scoring Alternatives in relation to the criteria** 4. **Determine the weights of the criteria.** 5. **Evaluating the alternatives.** 6. **Validating/Verify the Result** ### Step 1: Establishing the Decision Context - Identifying the aims of MCE decision-makers and other stakeholders. - Establishing a shared understanding of the problem situation. ### Step 2: Structuring the Decision Problem - Identifying objectives, alternatives, and criteria. | Should be: | | | |---|---|---| | S - Specific | Options that may contribute to the achievement of the objectives. | Uncontrolled variables that reflect performance in achieving objectives. | | M - Measurable | | | | A - Attainable | | | | R - Relevant | | | | T - Time-bound | | | ### Step 3: Scoring Alternatives in Relation to Each Criterion - **Linear Conversion:** Set the suitability values of criteria to a common scale to allow comparisons. Decision-makers decide which function should be used for each criterion. - **Fuzzy Membership:** Allows you to determine the likelihood that a location is suitable or not. Values range from 0 to 1 (or 1 to 100) with 0 being not likely and 1 being most likely. - **Negative Linear Relationship**: - Small Fuzzy Membership Type, Large Fuzzy Membership Type, Near Fuzzy Membership Type - **Reclassifying/Standardizing Criteria using a scale** ### Step 4: Determine The Weight of Each Criterion - **Ranking:** Rankings and ratings are usually converted to numerical weights on a scale from 0 to 1, with overall summation of 1 (normalization). - **Pairwise Comparison (Analytical Hierarchy Process (AHP)):** A matrix is constructed where each criterion is compared with the other criteria relative to its importance. - **Conservation:** 1 - **Profit:** 2 - **Flood:** 1 - **Erosion:** 1 - **Landscape:** 1/2 - **Conservation:** 1/2 - **Profit:** 1/2 - **Flood:** 1 - **Erosion:** 1/3 - **Landscape:** 1/4 - **Conservation:** 1/2 - **Profit:** 1 - **Flood:** 1/2 - **Erosion:** 1/2 - **Landscape:** 1/3 - **Conservation:** 1 - **Profit:** 2 - **Flood:** 1 - **Erosion:** 1 - **Landscape:** 1/2 - **Conservation:** 1 - **Profit:** 4 - **Flood:** 3 - **Erosin:** 2 - **Landscape:** 1 - **Weight:** - Conservation: 0.2 - Profit: 0.2 - Flood: 0.125 - Erosion: 0.18 - Landscape: 0.19 - Conservation: 0.21 - Profit: 0.1 - Flood: 0.125 - Erosion: 0.09 - Landscape: 0.13 - Conservation: 0,1 - Profit: 0,1 - Flood: 0,125 - Erosion: 0,18 - Landscape: 0,13 - Conservation: 0,2 - Profit: 0,2 - Flood: 0,125 - Erosion: 0,18 - Landscape: 0.19 - Conservation: 0.4 - Profit: 0.4 - Flood: 0.375 - Erosion: 0.36 - Landscape: 0.39 - SUM: 1 - **Weight of Priority is also called Normalized Principal Eigen Vector.** - To normalize the values, divide the cell value by its column total. - To calculate the weight, determine the mean value of the rows. - **Consistency Ratio (CR):** Measures how consistent the judgments have been relative to large samples of purely random judgments. - **CR = Consistency Index (CI) / Random Consistency Index (RI)** - **CI = (Amax-n) / (n-1)** - $$A_{max}$$ is the maximum value from each normalized row of the matrix. - n is the number of factors. - Sum of each column from the original table. - Sum of the columns from the normalized table. - **Benefits of Using AHP:** - More structured approach for measuring suitability by breaking down the problem into hierarchical criteria. - A more systematic, in-depth analysis of the factors. - AHP allows the participation of experts and stakeholders in providing input and the incorporation of both quantitative and qualitative criteria. ### Step 5: Evaluating the Alternatives - **Non-Weighted MCE (binary).** - Criteria are classified as good (1) or bad (0). - Determined by multiplying the three binary preference maps, so values OxOxO and Ox1x1 are the same (AND). - Adding instead of multiplying can show areas of the map on a scale that fit some of the criteria (OR). - **Weighted Linear Combination (WLC):** Most commonly used decision rule. - **Other Methods:** - Non-weighted geometrical mean summation of the factors. - Ordered weight average (OWA). ### Step 6: Validate/Verify The Result - Ground truth verification (ex: conduct a field survey to verify sample areas). ### Sensitivity Analysis - How do the following affect the results? - Altering the set of criteria (more/less) - Altering the weights of the factors. - Is the result reasonable? - Does the result reflect reality? - Etc. ### GIS-Based MCDA (Multicriteria Decision Analysis) - **Criterion Maps:** - 96000 - 80000 - 100002 - **Standardisation:** - 0.85 - 1 - 0.73 - **Standardised Criterion Maps:** - 0.80 - 1 - 0.50 - 4 - 4 - 4 - **Weighting:** - 0.667 - 0.2 - 0.125 - 0.667 - 0.25 - 0.125 - **Weighted Standardised Criterion Maps:** - 0.556 - 0.85 - 0.1485 - 0.083 - 0.50 - 0.33 - 0.028 - 1 - 0.1042 - **Summation via overlay:** - 0.839 - 0.82 - 0.77 - **Overall Score Map** - **Reclassify** - **Suitability Map** ### Remarks - GIS-based MCE is good for complex scenarios. - Subjectivity in choosing the criteria and defining the weights of each factor. - KISS approach (Keep it simple and stupid!) - Models are to be interpreted with caution. - Validation increases the credibility of the results. ## GIS Packages - **Teelset** has decision support modules: - Fuzzy (used to standardize factors). ­­ - Weight (used to calculate the AHP weights). ­ - MCE (for the actual evaluation) and others. ­­ - The MCE process can also be done in ARCGIS (model builder) although it doesn't have the standardization functionalities like Teerset. - Python implementation. ## 3D GIS - 3D GIS maps illustrate objects in greater detail by adding another dimension (Z). - Even though Z-values are most often real-world elevation values, there's no rule that enforces this methodology. - **3D GIS Data** - Feature Data: - **DEM (Digital Elevation Model):** Raster representation of a continuous surface. Each cell of the raster has an elevation value. Analysis tools can be run on DEMs to produce new surfaces such as slope and aspect. - Surface Data: - **Raster surface Elevation DEM Model:** - **Contour Lines:** Familiar way of representing surfaces on maps. ­ - **Contour:** A line through all contiguous points with equal height (of other) values. - By following the line of a particular contour, you can identify which locations have the same value. - By looking at the spacing of adjacent contours, you can gain a general impression of the gradation of values. - **Raster Surface section> 3D Analyst Tool Box> Contour TOOL** ## TINs (Triangular Irregular Networks) - Form of vector-based digital geographic data. - Are constructed by triangulating a set of vertices (points). The vertices are connected with a series of edges to form a network of triangles. - **Methods of interpolation to form the triangles:** Delaunay Triangulation; distance ordering. (Supported by ARCGIS.) - Less widely available, more expensive, and less efficient than raster surface models. - ARCGIS can create and store 4 types of surface models: raster, TIN, terrain data sets, and LAS 2.5D data. (Functional surfaces.) - It's continuous and all locations on the surface only have 1 elevation. - **True 3D data:** Solid model surfaces can store more than one Z-value per (n,y) position. ## You Can Create a TIN Surface From: - Surface source measurements. - Features (points, lines, polygons) that contain elevation information. - Contour lines. - Another functional surface (raster, terrain dataset, LAS dataset). - **TOOL:** Create TIN. (3D Analyst Extension > TIN to Raster Tool) - Creates TIN from points, lines, and polygons. ## Show Elevation Data of TIN in 3D: - Elevation sources: - Ground: World Elevation3D/ Terrain 3D - Properties: Data source: Your TIN file. ## Derive New Surfaces From the DEM - **Create Aspect Surface:** - Aspect tool from 3D or Spatial Analyst. - Aspect is the 3D orientation of the slope. - **Create Slope Surface:** - Slope tool from Spatial Analyst. - Slope is the degree of incline of a surface (in %). - **Create a Hillshade:** - Hillshade tool. - Visualize terrain as shaded relief, illuminating it with a hypothetical light source. - Based on slope and aspect. - **Hillshade TOOL:** - Azimuth (where in the sky the sun is) and altitude (height) values. - Angle of the sun in the horizon. ## Visibility Analysis - **Viewshed:** Determines the raster surface locations (cells) visible to a set of observer features. - **Useful for when you want to know how visible objects might be from a certain location.** - **Viewshed tool (Spatial Analyst):** Draw your own interest points. ## WebGIS - **What is WebGIS?** - WebGIS is a pattern or architectural approach for implementing a modern GIS. - It's powered by web services - standard devices that deliver data and capabilities and connect components. - **WebGIS can be implemented** - In the cloud. - On-premises. - Hybrid. - **WebGIS is a transformation of GIS that brings analytics to spatial data in a way that wasn't possible before.** - Previously, spatial data had to be processed, modified and extracted to answer a predetermined set of questions. - Now, the data is transformed into web maps or services that are mashed up with different layers into a web GIS which provides the flexibility to answer any possible question. - The data is ready and waiting to dynamically answer questions. It no longer needs to be processed for each individual question. - WebGIS is a much more flexible and agile workflow. - **WebGIS brings GIS into the hands of the people.** - **What is a portal?** - Provides a framework for sharing and using maps, apps, and data. - Enables wider use and access to GIS data serving GIS professionals, knowledge workers, decision makers, and developers as well as public access. - Can be hosted in the cloud or on-premises. ## WebGIS Options (web.tools) - **Commercial:** - ArcGIS Enterprise - Geomedia Web Map / ERDAS Apollo - Super Map (Edge GIS Server) - **FOSS (Free Open Source Software):** - Geoserver - Mapserver - QGIS Server ## OGC (Open Geospatial Consortium) Standards - WMS (Web Map Service) - WFS (Web Feature Service) - WES (Web Coverage Service) - KML (Keyhole Markup Language) - WFST (Transactional Web Feature Service) ## GIS In The Cloud - Carto DB - GIS cloud - Map 2 Net - Mapbox - Mango Map - ArcGIS Online - **Google maps:** Don't allow GIS spatial analysis. - **Bing maps:** Don't allow GIS spatial analysis. - **Google Earth:** Don't allow GIS spatial analysis. - **Google My Maps:** Allows you to share your data. ==End of OCR==