Intermediate GIS Final

Euclidean Distance

* Simplest expression of distance in GIS
* d = SqRt{(x2-x1)^2 + (y2-y1)^2}
* Pythagoras’s Theorem and the straight-line distance between two points on a plane

Euclidean Distance Tools

* Euclidean Distance: gives the distance from each cell in the raster to the closest source (What is the distance to the closest town?)
* Euclidean Direction: gives the direction from each cell to the closest source (What is the direction to the closest town?)
* Euclidean Allocation: identities the cells that are to be allocated to a source based on closest proximity (What is the closest town?)

Static Modeling

The series of steps required to achieve some final result, set of linear steps or multiple linear steps towards an end (Siting of a new school/store)

Dynamic Modeling

The series of steps required to achieve some final result but has additional parameters requiring several iterations of the model, some inputs will change based on different scenarios (disease outbreak)

To problem solve: Simplified Modeling

* Identify the problem
* Breakdown (simplify) the problem
* Organize the data required to solve the problem
* Develop a clear and logical flowchart using well defined operations
* Run the model and modify it if necessary

Multi-Criteria Evaluation Steps

1. Set the goal/define the problem
2. Determine the criteria (factors/constraints)
3. Normalize the factors/criterion scores
4. Determine the weight of each factor
5. Aggregate the criteria
6. Validate/verify the result

What is remote sensing?

* Information is derived from measurements of the amount of electromagnetic radiation reflected, emitted, or scattered from individual objects
* Includes satellite imagery, aerial photos (taken from aircraft/drones, helicopters, balloons, masts)
* Most RS is performed from orbital (spacecraft) or sub-orbital (aircraft) platforms using instruments which measure electromagnetic radiation reflected or emitted form the terrain

Passive system of remote sensing

* Detect natural radiation that is emitted or reflected by the object or surrounding area being observed
* Utilize reflected electromagnetic energy from the sun (most common source of radiation measured by passive sensors)
* Ex: Regular digital photography, and many commercial satellites

Active system of remote sensing

* The sensor emits energy in order to scan the area of interest, and then it measures the radiation that is reflected or backscattered from the target area
* Emits and detects other energy sources such as radar or laser
* RADAR is an example, where the time delay between emission and return is measured, establishing the location, height, speed and direction of an object

Electromagnetic Spectrum

* Most common types of remote sensing are based on reflected electromagnetic energy
* Radiant energy is classified by wavelength and organized into a chart known as the electromagnetic spectrum
* Wavelength is the distance between the crests of the waves in a beam of light
* Visible spectrum has a wavelength between .4 and .7 micrometers
* Electromagnetic energy is either reflected, absorbed, or transmitted

Visible spectrum and color

* The perceived color of light has a distinctive wavelength
* Most colors can be found by mixing the bas colors red, green, blue, with varying levels of brightness
* The perceived color of an object is the result of the wavelengths that are reflected

Spectral Reflectance Curves

* Each entity has unique spectral signatures and they differ with each band of information
* Each feature that we see has its own spectral reflectance curve which are defined by the varying percent of reflectance
* In RS, one must understand the reflectance nature of an object if it is going to be identified on an image. Graphs of spectral reflectance curves help us better understand the reflectance nature of an object
* A surface feature’s color can be characterized by the percentage of incoming electromagnetic energy it reflects at each wavelength across the electromagnetic spectrum. This is its spectral reflectance curve or “spectral signature”; it is an unchanging property of the material

Atmospheric Opacity

* Energy emitted by the sun broad spectrum, energy increases rapidly to a maximum around visible
* Only parts of the electromagnetic spectrum are transmitted through the atmosphere

Remote Sensor

* Stores the visible and non-visible electromagnetic spectrum in different bands
* With the level of each reflectance store as a digital #
* Combining red, green, and blue bands can create a true color composite

Image Bands

* Sensors can pick up on more than red, green, and blue light, like near-infrared which is not visible to the human eye
* Blue has the shortest wavelength (450-515 nm), green is in the middle (525-605nm), red has the longest wavelength (640-690 nm) apart form NIR (750-900 nm)

Band Density

* Broad Band: Few broad (wide) bands chosen for specific end-user purposes
* Hyper Spectral: Many very narrow bands, can be manipulated for multiple purposes, but data intensive

Band Composites

Bands can be mixed to create false-color composites or to enhance other image features

Band Combinations

True-color 321

* Red (3), green (2), blue (1)
* Band 3 detects chlorophyll absorption in vegetation
* Band 2 detects the green reflectance from vegetation
* Band 1 is more suited for penetration in water, in clear water this can be some 25 meters

False-color

* Near infrared, red, green
* Makes vegetation appear as red-tones, brighter reds indicting more the growling vegetation
* Soils with no or sparse vegetation range from white (sand, salt) to greens or browns depending on moisture and organic matter content
* Water appears blue; clear water will be dark blue to black while shallow waters or waters with high sediment concentrations are lighter blue
* Urban areas will appear blue towards gray

Spectral Reflectance of Vegetation

Chlorophyll strongly absorbs radiation in the red and blue wavelengths but reflect green wavelength

Normalized Difference Vegetation Index (NDVI)

* Commonly used to measure and monitor plant growth, vegetation cover, and biomass production from multispectral satellite data
* Healthy vegetation has low red-light reflectance and high near-infrared reflectance
* NDVI = (NIR-red)/(NIR+red) or NDVI=\[(B4)-(B3)\]/\[(B4)+(B3)\]

Chlorophyll-a

Can be used as a proxy for phytoplankton and thus is an essential water quality parameter. The presence of phytoplankton in the ocean causes selective absorption of light by chlorophyll-a pigment resulting in change of the ocean color that can be identified by ocean color remote sensing

NDVI Problem

* NDVI saturates as leaf area increases
* NDVI might be a poor measure of plant growth in very dense ecosystems

Spatial Resolution (Data Resolution)

Pixel cell size (Landsat: 30m, IKONOS: 4m), image extent (swath width)

Spectral Band Resolution (Data Resolution)

The number and size of spectral regions the sensor records data in, e.g. blue, green, red, NIR, thermal etc

Temporal Resolution (Data Resolution)

How often a satellite collects data for one location

\
Repeat Coverage

* AHVRR (Twice daily- AM and PM)
* MODIS (2 times daily)
* Landsat (every 16 days)
* Spot (Every 26 days)

Radiometric Resolution (Data Resolution)

Sensitivity of the detectors to small differences in electromagnetic energy

Coverage

Individual images with an assigned “row” and “path” can be downloaded and mosaicked to create larger images

What satellite imagery would I use for disaster management?

Geoeye

What would I use for thermal imaging with UAV?

Dragonfly

What would I use for land cover classification?

MODIS

Classification

* Refers to the task of extracting information classes from a multiband raster image
* Classes are often easily distinguishable by plotting their level of reflectance for 2 bands, since each reflects different amounts of energy in each band
* Resulting raster can be used to create thematic maps
* Two types of classification: supervised and unsupervised
* Process can be done on more than 2 dimensions

Sources of Classification Information

1. Multiple Bands: Capture key parts of visible and non-visible spectrum
2. Reflection Intensity: Digital Number (DN) portrays energy intensity for each band
3. Repeat coverage: Sites often revisited many times in a year, historical record within and between years

Unsupervised Classification

* Land-classes determined by algorithm without data input form user
* Uses no ground truth data. Instead identifies separable data clusters. Where the number of clusters is often determined by the user. Class labels must be determined by the user after the clustering analysis is done

Supervised Classification

Land-classes determined by user and passed to algorithm for classification of remainder of image

ISOData Clusters

* An iterative search adapts clusters to minimize within cluster variability with the function for sum of square distances
* The Iso Cluster Unsupervised Classification tool automatically finds the clusters in an image and outputs a classified image.
* Uses an isodata clustering algorithm to determine the characteristics of the natural groupings of cells in multidimensional attribute space and stores the results in an output ASCII signature file.

Supervised Classification Steps

1. Identify land-use of training sites (multiple observations of each land class to represent variability)
2. Pull satellite band data to each training site
3. Train classification rule
4. Apply rule to entire image

Minimum Distance (Supervised Classification Techniques)

* Minimizes the distance between any point and the centroid of a defined class
* This maximizes the similarity of points within a class but class means determined by user

Maximum Likelihood (Assigning classes to new observations)

Given the distribution of known points and user defined classes, assuming a normal distribution, for unknown points assign the class with the greatest probability

Multiday Composite

Helps avoid effects of cloud cover and haze

LandSat

* Uses: Regional scale dynamics: land, cloud, oceans. Series of satellites
* Coverage: every 16 days
* Nine spectral bands

MODIS

* Uses: Large scale dynamics: land, cloud, oceans. Combines information from 2 satellites Aqua (PM) and Terra (AM)
* Coverage: Global every 1-2 days
* Resolution: 250m (bands 1-2), 500m (bands 3-7), 1000m (bands 8-36)
* Swath width: 2330 km
* Launched: Y2000
* Is being retired

Spot

* Uses: Local high resolution land-cover, change, forestry
* Responsive tasking
* Coverage: Daily possible with constellation
* Resolution: Spot 6-7, 8m (Bands 1-4), 1.5m Panchromatic
* Swath Width: 60 km
* Launched: multiple since 1986, 2014 new

Planet Labs: Microsatellites

* Constellation of 100 satellites
* Daily coverage for globe since 2015
* Small satellites “doves”: 3-5 meter resolution
* Larger satellites (SkySat): .5 meter resolution
* RGB and NIR bands

Data Access

* Classified Land Cover Products: National Land Cover Database, MODIS Land Cover
* Unclassified data products: Aerial, LIDAR, Radar, AVHRR, MODIS, Aster, Landsat, global land survey

Data Abundance and Inequity

* 2.5 quintillion bytes of data created every day
* 947 Earth observation satellites in operation
* 8 countries operate 86% of all satellites in orbit
* 40+ commercial satellite companies collecting images of a “digital earth”
* 1/3 of the global population lack reliable online access

Interpolation

* Estimating the value of properties at unsampled sites within the area covered by existing observations
* Widely used for: climate data and predictive stats
* Data should be: point files, interval or ratio
* Keep in mind: Tobler’s first law, must be spatially autocorrelated, generally only works for continuous variables

When is this statistical tool appropriate?

* When the basic assumptions are satisfied: Data is spatially dependent and continuous within the given bounds
* Examples: elevation, temperature, rainfall, chemical dispersion

Categories of Interpolation

* Deterministic: IDW, Natural Neighbor, Trend, Spline
* Geostatistical: Kriging
* Thiessen Polygons (technically a proximity tool)

Spatial input data are usually

* Stratified (Regularly spaced observations)
* Patchy (Clustered observations at specific locations)
* Random (Unevenly located)
* Adaptive (this method for estimating feature variation is used to locate sample sites in areas of high variation)

Thiessen Polygons

* This is an exact method of interpolation that assumes that the values of un-sampled areas are equal to the value of the closest sampled point
* This method assumes the values of un-sampled area are equal to the value of the closest sampled point
* Mathematically defined by the perpendicular bisectors of the lines between all points

Triangulated Irregular Network

* Used to construct Digital Elevation models
* Adjacent data points are connected by lines to form a network of irregular triangles

Autocorrelation and Distance decay

* Tobler’s first law of geography: everything is related to everything else but near things are more related than distant things. Use this to estimate unknown values by weighting nearer objects higher than distant ones (only works w/ + spatial autocorrelation)
* Spatial Trends: A locations attributes can often be predicted based on their proximity to other spatial features (Ex: House property values typically decline with distance to good schools)

Inverse Distance Weighting (IDW)

* Determines cell values using a linearly weighted combination of a set of sample points. The weight is a function of inverse distance
* Values determined as the weighted average of nearby points: weights usually calculated using inverse distance weighting, interpolated value is an average over the observed values
* Equation found on slides

Distance to the power of x

* n = power of 1: Distant points strongly influence neighbors
* n = power of 2: Moderate influence
* n = power of 3: Distant points weakly influence neighbors
* Lower the power, the smoother the data

Inverse Distance Weighting Interpolation

* Works best for evenly spaced observations
* Doesn’t fit known data perfectly (Because the estimated are averages, the resulting surface will not pass through the sample points)
* Cannot make estimates outside of max and min observed values

Problems with IDW

* In IDW we arbitrarily determine the weight placed on each point using distance

Kriging

* Geostatistical Method: Focuses on the statistical relationships among the measured points, Uses the “ideal” fit to determine weights, “Ideal” fit is determined by a variogram
* Assumptions: the distance or direction between sample points reflects a spatial correlation that can be used to explain unknown values for the surface

The Kriging Process

1. It creates a semi-variogram cloud (measure of dissimilarity) to estimate the statistical dependence (spatial autocorrelation)
2. The shape and degree of autocorrelation is represented by a line of best fit (non-linear line)
3. It predicts the unknown values based on the line of best fit (making a prediction)

Spatial weighting in Kriging

* Choose some point
* Build a set of distance bands around it
* Calculate dissimilarly ( Semivariogram(distanceh) = 0.5 \* average((valuei – valuej)2)

What is dissimilarity measured by?

Semi-variance

The Variogram

* The line of “ideal fit”
* It traces the increasing decline and spatial autocorrelation (from one point to any other) as the distance between them increases

What kind of relationship does the variogram have with spatial weights?

Inverse relationship

How are we measuring dissimilarity?

* Semivariance = Dissimilarity
* Semivariance sums the squared difference (xi-xj)^2 between pairs of xi and xj that are specified distance h
* Lower values of Svh = greater correlation between locations xi and xj that are h distance apart

The difference between IDW and Kriging

* IDW: weights depend solely on the distance to the prediction location
* Kriging: weights are based on the distance between the measured points and the predication location AND the overall spatial arrangement of the measured points (Spatial Autocorrelation)

Common Patterns of Web GIS Use

* Mapping and visualization
* Data management: State GIS Portals like DC Open Data
* Field mobility: Logistics, confirming delivery of package with geotag image etc. 
* Monitoring
* Analytics: Analytics portals – eg wildfire probability estimates
* Design and planning: Commercial property development, testing layouts, simulating neighborhood
* Decision support: COVID dashboards
* Constituent engagement: community mapping, 611 app geotag potholes etc.
* Sharing and collaboration: Share model outputs with end users.

Basic Architecture of Web GIS

* Database server
* GIS server
* Web server
* Client
* Client requests information through the internet and internet communicates with the servers to retrieve the response

Content items

* Data, is published to
* Web layers, which is added to
* Web maps and web scenes, which are configured and used in (tools are used to help configure)
* Web apps, mobile apps, and desktop apps

Web layer - Tiling

* Vector tiles are packets of geographic data, packaged into pre-defined roughly-square shaped “tiles” for transfer over the web
* Lower (0) tiles have much less detailed information
* Only data within the current map view, and at the current zoom level need to be transferred
* Greatly reduces the amount of data that is transferred on web

Basic components of a Web GIS app

Basemaps + Hosted Operational layers + Tools

Basemap layers

* Provide reference or context
* ArcGIS Online includes a gallery of 2D and 3D basemaps
* ArcGIS Online can use your map services as the basemaps

Hosted Operational layers

* Thematic representations of your data
* Can be created or discovered
* Many types: Feature, tile layer, map image layer, 3D scene layer, image layer, KML, GeoRSS

Living Atlas of the World

Maps and layers from Esri and thousands of contributors

Tools

* Perform tasks beyond mapping
* Examples: routing, geocoding, printing, querying, summary

Technology evolution and trends

* From one-way to two-way information flow
* Portal technology is becoming essential
* Cloud GIS accepted as the primary way to deliver GIS functions
* Mobile is becoming the pervasive Web GIS client platform
* Map visualization goes from 2D to 3D, to virtual reality and augmented reality
* Data source goes from static to real-time and spatiotemporal big data
* Web GIS becomes smarter and more intelligent

Style your layers: Smart mapping

* Workflows analyze your data and suggest the best way to represent your data
* Smart defaults take the guesswork out of setting up many of the map properties
* Based on the basemap you select, smart mapping automatically suggests and coordinates colors and other map styling
* Suggested visible ranges allow you to see your data at sensible scales

Smart mapping: Many types of styles

* Heat map
* Color map
* Size map
* Point map
* Color and size map
* Time map
* Arcade expression map
* Predominance map

Web map - Zoom level styling

* The zoom level is an integer that specifies the level of detail to display on the map, with larger integers resulting in greater levels of detail
* Designers choose what information is provided at each zoom level

Pop-ups: Windows to geospatial information

* Displays when someone clicks on the map
* Supports the following content types: Attribute fields (aliases and values), custom-formatted text, attachments, images and videos, charts, links

ArcGIS Arcade

* Portable, lightweight, and secure expression language
* Build expressions without having to alter the underlying data
* Expressions can be used to style layers and in pop-ups
* Expressions created in web maps are respected in most ArcGIS web and mobile apps

Attachment Viewer

* Present attachments, including geotagged photos, in an intuitive browsing experience
* Has two types of layouts: map-centric and attachment-centric
* Works well in desktop and mobile

Stages in Data Collection Projects

* Planning: includes establishing user requirements, garnering resources, and developing a project plan – what do you need to achieve your research objective (shopping list)
* Preparation: involves obtaining data, redrafting poor-quality map sources, editing scanned map images, removing noise, setting up appropriate GIS hardware and software systems to accept data. Getting ready to ingest them into the software environment
* Digitizing and transfer: are the stages where the majority of the effort will be expended. Capturing geometry from scanned layers or imagery, or importing tables, databases
* Editing and improvement: covers many techniques designed to validate data (checking the topology), as well as correct errors and improve quality, cleaning up and classifying attribute tables
* Evaluation: is the process of identifying project successes and failures – usually this is done after you’ve performed your analysis, or ran your model (unusual finding, or inconsistent finding –could this be due to data quality, or format)

Data Collection

Two main methods: Primary data collection (direct measurement) and Secondary data collection (derivation from other sources)

Vector Data Capture

Two methods for collecting locations in the field: Traditional ground surveying methods and Global navigation satellite systems (GNSS)

Position by ranging

* GPS receivers calculate distances to satellites as a function of the amount of time it takes for satellites’ signals to reach the ground
* To make such a calculation, the receiver must be able to tell precisely when the signal was transmitted and when it was received. The satellites are equipped with extremely accurate atomic clocks, so the timing of transmissions is always known. Receivers contain cheaper clocks, which tend to be sources of measurement error
* 2D example: if you know you are a certain distance from a known position, you could be anywhere edge of the circle. The more known locations you add to the equation, the closer you are to determining your position. With three sets of known points and distances you are able to determine a new known location

Traditional Ground Surveying Methods

* Encompass traditional surveying through angle and distance measurements
* Coordinate surveying uses measurements of distances and angles to estimate coordinate locations
* Land surveyors measure horizontal positions in geographic or plane coordinate systems relative to previously surveyed positions called control points
* First position is measured relative to the stars

Surveying

* The art and science of measuring the surface of the earth and its features
* Geodetic surveys: take into account the true shape of the earth
* Plane surveys: treat the earth as a flat surface
* Horizontal surveys determine the position of features on the ground
* Vertical surveys determine the elevation or heights of features

Measuring angles

* Bearings: an angle less than 90 degrees within a quadrant defined by the cardinal directions
* Azimuths: an angle between 0 degrees and 360 degrees measured clockwise from North

Horizontal Datums

A reference surface with two specific components

\
* Ellipsoid: 3D model complete with Geographical coordinate system
* Terrestrial reference frame: a physical network of precisely measured points

Control Networks

A series of well-spaced and interconnected markers in the ground which have accurately determined positions, or coordinates, and elevations for:

* Mapping our natural resources
* Locating international and provincial boundaries
* Navigation over land or water
* Producing topographic maps
* Planning and building highways, pipelines and other major engineering projects
* Mining and funneling operations
* Building construction

Horizontal Surveys

* Traverse (open or closed)
* Triangulation
* Trilateration

Instruments for Horizontal Surveying

* Theodolite surveys: measures the angles, and the distances are measured with either a steel measuring tape or, more commonly, an electronic distance meter (EDM).
* Total Station surveys: is a theodolite and EDM wrapped up into one instrument. The measurements (angle and distance) are made electronically, and stored digitally. Survey measurements are carried out easier, faster and more accurately. Back in the office, software processes the data and produces a plan automatically
* GPS
* LIDAR: (LIght Detection And Ranging) - Lasers on board an aircraft 'scan' the ground surface calculating distances. Using the distances, the angle of measurement and the location of the LIDAR scanner itself, positions and elevations are calculated.

Leveling

* A vertical survey establishes an elevation relative to a reference surface
* Methods: Differential leveling and trigonometric leveling

Differential leveling

this method of establishing elevations uses a tripod mounted telescope, called a level, which is aligned at right angles to the direction of gravity. The elevation is determined by measuring the difference in readings taken on two graduated rods similar to very large rulers. Points with known elevations are referred to as "benchmarks". If we know the elevation of one point, we can "transfer" the elevation for a new point by applying the difference in elevation between the known (or assumed) elevation and the new point.

Trigonometric leveling

This form of transferring elevations uses the mathematics of trigonometry. Rather than using a level, we can use the same theodolite we used for our horizontal surveys. The theodolite is set up over a point of known elevation (benchmark) and the height of the centre of the telescope above the reference point is carefully measured and added to the known height to produce an accurate HI, or height of instrument.

Cadastral Surveys

Deals mainly with the establishment or relocation of land boundaries

\

1. To acquire the data needed to write a legal description of a specific parcel of land
2. To re-establish the boundaries of a parcel of land for which a survey has previously been done
3. To subdivide a parcel of land into various pieces according to a specific plan showing the predetermined size, shape and location for each parcel

Importance of Time

Since the speed of the signal is known, to solve for distance from one point to another all that is needed is deltaT. In two-way ranging, the EDM clock is perfectly synchronized with itself, and can solve easily for deltaT

Two-Way Ranging (ground surveying)

The EDM records the elapsed time between the wave's emission and its return from the reflector. It then calculates distance as a function of the elapsed time

\
Distance = deltaT (c)/2

How GPS Works

1. Tracking stations use radio signals to determine orbits of GPS satellites.
2. Command center transmits orbital data, time corrections, and location of other satellites in the GPS constellation.
3. GPS satellites simultaneously transmit synchronized time and orbital data to Earth.
4. GPS receivers compute location using orbital data and the difference in arrival times of the signals of at least 4 satellites.

Orbital Planes of the GPS System

* Arranged to ensure that at least 4 satellites are “in view” at any given time
* 3 satellites are needed by the receivers to determine position, while the fourth enhances the measurement and provides the ability to calculate elevation
* Since four satellites must be visible from any point on the planet and the satellites are arranged into six orbital planes, the minimum number of satellites needed to provide full coverage at any location on Earth is 24.

Time and Navigation

* Atomic clocks in GPS satellites keep time to within three nanoseconds—three-billionths of a second. Position accuracy depends on the receiver. Most handheld GPS receivers are accurate to about 10 to 20 meters (33 to 66 feet).
* GPS Accuracy: Both military and civilian users can obtain higher accuracy by using a second GPS unit at a fixed nearby location—a method called Differential GPS. In this way, positions can be determined with an accuracy better than 1 centimeter (less than half an inch). For military users, additional encrypted signals can provide high accuracy.
* Synchronizing GPS: All GPS satellites must transmit their data signals at the exact same time, so precise synchronization is essential. Their signals are monitored constantly and adjusted as needed. The GPS Operations Center at Schriever Air Force Base in Colorado Springs, Colorado, controls the constellation of satellites that provides navigation data to military and civilian users worldwide
* Code in GPS Signals: Signals sent by GPS satellites are complicated. What we learn from these signals depends on the type of equipment we use to receive them. A military pilot needs more information than someone using GPS in a car.
* More Accuracy for Professional Surveys: Specialized GPS equipment was developed for high-precision surveys during the 1990s. Differential GPS uses two sources of GPS data to provide high accuracy. Surveyors used this type of system to accurately map infrastructure and survey large sites.