Published on February 28, 2014
Sample Works Ravi R Sadasivuni www.gri.msstate.edu
Table of Contents Gravity Models 1. Economic and Retail applications in the Region Using GIS 2. Convection-Diffusion Based Spatial Interaction (Gravity) Model to Predict Anthropogenically-Initiated Wildfire Risk Other Models 1. A Transportation Corridor Case Study: A Multi-criteria Decision Analysis (MCDA) 2. Digitization & Topology Corrections 3. Morphometry & Landform Classification www.gri.msstate.edu
Economic and Retail applications in the Region Using GIS www.gri.msstate.edu
Background Specific Gravity • Total gravity surface derived and then a specific gravity (capture rate) calculated for site in question • The total number of consumers captured are determined by population density. • Distance can be measured as a crow flies, along a transportation network or in terms of estimated travel time. www.gri.msstate.edu
Economic Impact Modeling the economic impact of adding new competitive locations displayed from Red (High) to Low (Blue) Hypothetical Locations Potential Trade region www.gri.msstate.edu
Trade influence in the Southern Mississippi High Low www.gri.msstate.edu the site and situation of the center influential to economy with respect to factors such as the type and age of center, proximity of competitors, accessibility, visibility, clientele, topography, crime, and socioeconomic factors. Thus, it is essential to account for market structure and the types of products or transportation for proper use of gravity models.
Previous Prediction Models United States Forest Service models use correlations between the historical observational and geographical data, climatic and vegetation components to obtain fire risk potential indices There are few models which account both fuel loads and ignition sources to predict wildfire potential Road Density and road buffers tells about the activity along corridors but not the activities (from traffic vol) and its origin from the cities (Cooke and Grala, 2010; Gilreath, 2006) Gravity based model gives a GLOBAL measure of propulsiveness and a measure of destination attractiveness as the flow of activity between different zones, but does not consider the transportation corridors (Sadasivuni et al., 2013). Most wildfire ignition distribution models fail to define the diffusion direction, while diffusion seldom occurs across the Euclidean distance, but along the transportation corridors (Ayeni 1979) www.gri.msstate.edu
Gravity Model A Large city is more likely to attract an individual than a smaller city Newton’s formula is modified with physical mass for population size PiPj and a friction (cost) is inverse power function of distance f(cij)=dij-λ Pij = PiPj / dij-λ where λ is a parameter of the function dij is the distance between i and j (Sadasivuni et al. 2013, “Ecological Modeling”) “Perhaps the best-known model of spatial interaction is the gravity model, so called because of its analogy with the Newtonian concept of gravity. The traditional gravity model is based on the hypothesis that the interaction between any two masses varies directly with the product of the masses and inversely with some measure of the spatial separation or cost between the two masses” Predicting human activity patterns of incendiary activity using diffusion and convection with Newton’s Gravity Model. www.gri.msstate.edu
Background: SIM Road Density as Human Ignition (Gilreath, 2005) Not Accurate for high risk zones, such as along roads www.gri.msstate.edu
Background Gravity Model www.gri.msstate.edu Shows good interaction among cities, aligned along the roads In the broadest sense, the movement of people/animals/commodity/capital or information across the geographic space are modeled with origin, destination and transportation routes (in enhanced method later) that solve the problems of dispersion through GIS decision analysis. Gravity Model previously applied for wildfire prediction Performs poorly than Road density model in Medium risk zones (outskirts of the city), and very low risk zones
Approach: Modeling Philosophy Anthropogenically initiated wildfire is directly related to the movement of people Movement of people are: general guided by the cities aligned with the transportation corridors anisotropically aligned along high traffic volume roads provide access points to the woods for people with incendiary motive Influencing factors grouped as global and local variables. Global variables are modeled using Gravity term Local variables are parameters of transportation corridors modeled as convective/diffusive fluxes www.gri.msstate.edu
Background: Wildfire Impact Fig. Human disturbance of wild vegetation of North America. Redrawn from Man and Earth’s Ecosystems, by Charles F. Bennett (1975) Complex dynamics of coupled human-natural systems Fig. Schematic diagram showing the interaction between human activities from population interaction, location and time and its effects on the ecosystem through land use and wildfire. www.gri.msstate.edu
Wildfire Ignition Flow Chart: CDM Modeling Details Wildfire Risk Anthropogenic Factors Fuel P-E Climate Ignition Local Global Population Intermodal Transportation Vegetation Gravity (P/d2) Seasons Damping () Roads Rails Additive Multiplicative Diffusion (√ ) Diffusion (√ ) www.gri.msstate.edu Traffic Vol Convection V Vt Lro
Wildfire Ignition: CDM Formulation Occurrence of fire () depends on fuel (f) and ignition potentials (i ) ( x, y, t ) f ( x, y, t ) i ( x, y) f varies in space (vegetative characteristics) and time (variation of the seasons) i assumed to be function of space only and depends on, proximity (D) to city or population area (P); proximity (dro/ra) to railroads (ra) and roads (ro); traffic volume (Vt) Is a function of time, but not considered herein; length or density of the corridors (Lro/ra) www.gri.msstate.edu
Wildfire Ignition: CDM Formulation i B 1 LB V Vt Lro Convection along Roads 2 2 d ro / ra d Diffusionalong Corridors P D2 Gravity from Cities where, V is the convection velocity, is the diffusivity coefficients and is a gravity model damping term inside the city. (Sadasivuni et al., submitted to Ecological Modelling) Model couples the multi-criteria behavioral pattern in a single dynamic equation Unknown coefficients to be calibrated using observational data www.gri.msstate.edu
Observational data ± Study area - Southeastern Mississippi in Mississippi Primary roads - white lines (from MARIS) Railroads - black lines (from MARIS) Cities and its population (from MARIS) Wildfire incidences for years 1992-2009 (black symbols) (United States Forest Service ) Traffic volume in yellow symbols (gomdot.com) Black horizontal and vertical lines show the quadrats for the rail-road interaction analysis www.gri.msstate.edu
Observational Data Analysis : Fires vs Roads • Wildfire decreases with the increase of the distance from the roads • Historic wildfire events across the road buffers helped evaluate the ignition potential and calibration of ν (nu) which is 1/1250 through the curve fit of the observational data www.gri.msstate.edu
Observational Data Analysis : Fires vs Intermodal Transportation Roads and Rails run in parallel, so rail buffers alone cannot be analyzed. Wildfire events per unit length of the rail roads were computed in the study region for the years 1992-2009. Quadrats with Inter-modal transportation corridors show 30% higher wildfire frequency www.gri.msstate.edu
Observational data Analysis: Fires With Cities Road density at the fire locations along with distance from the city center. Analysis is performed using the cities Hattiesburg, Laurel and Ellisville. Wildfire events increase along the city radius Peak frequency at the medium density road, consistent with Gilreath 2006. (Units: km / Sq. Km) www.gri.msstate.edu
Methodology Observational data : Fires vs Traffic Volume per Road Density www.gri.msstate.edu Anisotropy conditions c in the city potential field (gravity) due to high traffic volume in specified direction/roads as shown by the arrows
Methodology Analytic Model and Calibration: City Potential P D2 City Interaction is modeled following Gravity Model Damping function is derived from an exponential curve fit over the data www.gri.msstate.edu
Preliminary Results Analytic Model and Calibration: Transportation Corridors Ignition potentials due to the road and cities are somewhat independent of each other, thus can be treated separately. www.gri.msstate.edu
Preliminary Results Analytic Model and Calibration: Traffic Volume Traffic volume generates an anisotropic behavior of the city potential along the angular direction (c), c L1 V B B P D 1 V Vt Lro VTr ( c ) e 2 Tr ( c ) P D2 ( c ) 2 / 45 VTr ~ 1- 15 is traffic volume per road density Gaussian distribution along the traffic volume direction c www.gri.msstate.edu
Preliminary Results Analytic Model and Calibration: Total Potential Ignition potentials due to the road and cities are somewhat independent of each other, thus can be treated separately, therefore: Spatial distribution of the combined Road and City ignition potential for two different city and road connection patterns. www.gri.msstate.edu
Preliminary Results Depiction in Fortran Code (a) Depiction of the study region for the Fortran simulations. The cities are represented as filled circular regions. Roads (Black) and Rails (Brown) with straight line segments. www.gri.msstate.edu Test Study region around Hattiesburg area, 80 km x 50 km (b) The combined wildfire ignition potential is tested using an inhouse Fortran code Domain discretized using 1001x1001 points, i.e., 85m x51m resolution.
Preliminary Results City Interaction and Traffic Volume SIM diffusion along the inter-modal transportation corridors www.gri.msstate.edu The relationship between transportation corridors and spatial interaction Petal side of the Hattiesburg show more interaction behavior
Intermodal Transportation Corridors and City Interaction High fire ignition potential distribution showing anisotropy conditions of traffic volume High fire ignition potential distribution coincident with more fires in broken ellipses The high spikes signify the high activity areas www.gri.msstate.edu
Preliminary Results Quintile analysis of Fire Distribution Across Risk Zones Observed fires (or fire frequency) distributed in the quintile bins From the figure roads have high number of fires in the high risk zone and almost equal number in Medium and very high risk zones Similarly, the observed fires are highest in very high risk zone and next levels holding high fires are medium Combined model i.e. ‘total’ (Black and high risk zones dotted) performs better for wildfire frequency predictions than the city and road models alone www.gri.msstate.edu
Results Using fluid-dynamics analogy (FMA) model Intermodal Transportation Corridors, City Interaction and Fuel to Predict Wildfire Risk www.gri.msstate.edu
Conclusion and Future Work Conclusion: Attempted an examination of the linkages among anthropogenic factors and wildfire behavior Interaction between transportation corridors with traffic volume and city interaction disentangling diverse relationships of factors to identify risk probability The distribution of the number of wildfire events decay exponentially as the distance from the road increases www.gri.msstate.edu
DIGITIZATION AND TOPOLOGY ERRORS FIX www.gri.msstate.edu
Fixing Geodatabase Topology Errors Present Work As a part of Landform analysis, I am fixing the topology errors on the appended of counties in Arkansas www.gri.msstate.edu
Morphometry & Landforms www.gri.msstate.edu
Ridges and Valleys and Flat Lands Flat Lands www.gri.msstate.edu
Various Morphometric Indices www.gri.msstate.edu
Zoom out View Appalachians Zoom in View See the curvatures of slope extracted 0 (Black) - indication of concavity or sub-horizontal terrain with isolated peaks 1 (White) - indication of convexity or sub-horizontal terrain with deep incision Values distributed between 0-1 indicating various forms like domes, cylindrical, toe slopes, mesas, basins etc www.gri.msstate.edu
Local Shape is a Function of Curvature Various Parameters: DEM Slope Profile Curvature Planar Curvature Tan Curvature SPI TWI www.gri.msstate.edu
Morphometric Curvatures Extracted From the DEM Profile Curvature General Curvature Planar Curvature www.gri.msstate.edu
Combination of Curvatures and Topographic Position Index (TPI) Landforms extracted 1=canyons, deeply incised streams 2=midslope drainages, shallow valleys 3= upland drainages, headwaters 4=u-shaped valleys 5=plains 6=open slopes 7=upper slopes, mesas 8=local ridges, hills in vallyes 9=midslope ridges, small hills in plains, 10=mountain tops, high ridges. www.gri.msstate.edu
A TRANSPORTATION CORRIDOR CASE STUDY FOR MULTI-CRITERIA DECISION ANALYSIS www.gri.msstate.edu
Past Research Work 1. Past Research Work • Involved in developing new and innovative approaches to address environmental impact assessment aspects in various transportation related projects. • Environmental Impact Analysis study was conducted for the proposed I-69 / I-269 corridor which is sponsored by the U.S. Department of Transportation – Research and Innovative Technology Application (DOT-RITA). • Developed quantitative approaches to reduce subjectivity in weighting schemes for site selection. www.gri.msstate.edu
I-269 location A study area map for the proposed project shows I269 location study alignments in Mississippi from I-55 to the Tennessee boundary. The study area is defined by a 2 mile buffer around the location study alternatives where remote sensing data collection (aerial imagery, digital stereo imagery) will focus covering about 180 square miles. www.gri.msstate.edu
From Ranking to Weights using AHP Ex: Slope All factors/criteria are normalized prior to combined analysis Pair-wise comparison: Relative rank of areas (slope >20%) higher than (slope 15-10%) (rank = 9 / rank = 5) 9 7 5 4 1 www.gri.msstate.edu 43
DEVELOPED METHODOLOGY FOR TRANSPORTATION CORRIDOR CASE STUDY USING MULTI-CRITERIA DECISION ANALYSIS Wetlands Note how the existing roads are reused and avoidance areas are considered www.gri.msstate.edu
AHP based Generated routes match the existing routes Note how the existing roads are reused and avoidance areas are considered www.gri.msstate.edu
I-269 study bed Cumulative Cost Surface Defining Initial Corridor www.gri.msstate.edu 32
3D View: The proposed project shows I-269 location study alignments in Mississippi from I-55 to the Tennessee boundary. www.gri.msstate.edu
... financial results; Data analytics is a ... East Coast location, transportation ... set of Transportation Investment Generating Economic ...
Economic Aspects of Optimized Works Planning ... for Economic Policy Analysis, ... of Roadway Transportation and Traffic Safety Research ...
... utilization drop is evident in most recent transportation industry analysis ... location, transportation ... results in efficient transportation ...
... Department of Marine and Transportation ... OTB Research Institute for ... Rotterdam.000 TEU were transported in this hinterland corridor of ...
How it works: 1. Register a free 1 month Trial Account. 2. Download as many books as you like (Personal use) 3. Cancel the membership at any time if not ...
... to promote road safety best practices and to disseminate research results ... on a real corridor use ... research and data analytics ...
... Analytics & Data Enrichment to improve the QUAliTy ... Textual Analysis in Accounting Research (2015 ... Optimized truckloading in private ...
A Domain Decomposition Approach for Solving Optimal Economic ... New Developments in Location Analysis ... Scoring the Level of Research Results and ...
A networking and brokerage event for the second h2020 transport call! ...
Profiles of institutions, research abstracts and research papers were provided by the respective Intergraph Registered Research Laboratory. ...