Please use this identifier to cite or link to this item:
|Title:||Testing maximum entropy analysis to define population distributions|
|Authors:||Lozar, Robert C.|
Tweddale, Scott A. (Scott Allen)
Ehlschlaeger, Charles R.
Baxter, Carey L.
Burkhalter, Jeffrey A.
|Keywords:||Cities and towns|
Human beings--Geographical distribution
Geographic information systems
Maximum entropy analysis
|Publisher:||Construction Engineering Research Laboratory (U.S.)|
Engineer Research and Development Center (U.S.)
|Series/Report no.:||Technical Report (Engineer Research and Development Center (U.S.)) ; no. ERDC/CERL TR-18-22|
|Abstract:||The Urban Security research project attempts to characterize sub-national populations and urban dynamics that impact indicators of urban security threats. This work used the MaxEnt (Maximum Entropy) program to de-fine the probable locations of human habitation for Chittagong Division, Bangladesh. This represents the likely first application of maximum entropy analysis to the delineation of human populations. The model was parameterized with a large number of known house locations and several predictor geographic information system (GIS) data layers that might influence the geographic distribution of human habitation. Analyses and tests done using MaxEnt indicated that the MaxEnt model effectively delineated the probability distribution of human habitation based on a series of inputs based on a series of inputs from three data sources: Landsat NaturalVue TM (Thematic Mapper) imagery, Shuttle Radar Topography Mission (SRTM) elevation, and OpenStreetMap roads. Results show that, by using remotely-sensed data and “presence only” samples of human habitation, it is possible to estimate the habitation distribution probability based on sound statistical procedures used in MaxEnt.|
|Gov't Doc #:||ERDC/CERL TR-18-22|
|Rights:||Approved for Public Release; Distribution is Unlimited|
|Size:||34 pages / 1.48 Mb|
|Types of Materials:||PDF/A|
|Appears in Collections:||Technical Report|