Population dislocation

Description

This analysis computes population dislocation based on a particular hazard. First, housing units, with detailed characteristics (tenure, household size, occupied, or vacant) are allocated to the address points (buildings). This is done by calling the Housing unit allocation analysis. After the housing units are allocated, the hazard event defined by calling Fragility and Hazard services would determine the value loss for each structure which would be the input for the dislocation calculation. The dislocation is calculated from four probabilities of dislocation based on a random beta distribution of the four damage factors presented by Bai et al. 2009. These four damage factors correspond to value loss. The sum of the four probabilities multiplied by the four probabilities of damage states is used as the probability for dislocation. Since the process to determine which households are dislocated is probabilistic an integer value being imported to seed the random number generator determines if a household dislocates.

Additionally, the Block Group characteristics, percentages of African-American and Hispanic population are taken into account.

The output is a CSV file with dislocated households and related variables.

Contributors

• Science: Walter Gillis Peacock, Nathanael Rosenheim

• Implementation: Nathanael Rosenheim, Yong Wook Kim, Gowtham Naraharisetty, Michal Ondrejcek, Chen Wang , and NCSA IN-CORE Dev Team

Related publications

• Lin, Y.S., Peacock, W.G., Lu, J.C., and Zhang, Y. (2008). Household dislocation algorithm 3: A logistic regression approach 08-05R. Retrieved from hrrc.arch.tamu.edu/publications/research reports/08-05R Dislocation Algorithm 3.pdf

• Rosenheim, N., Guidotti, R., Gardoni, P., and Peacock, W.G. (2019). Integration of detailed household and housing unit characteristic data with critical infrastructure for post-hazard resilience modeling. Sustainable and Resilient Infrastructure DOI: 10.1080/23789689.2019.1681821

• Bai, J., Hueste, M.B.D. and Gardoni, P. (2009), Probabilistic Assessment of Structural Damage due to Earthquakes for Buildings in Mid-America Journal of Structural Engineering 135(10) DOI: 10.1061/(ASCE)0733-9445(2009)135:10(1155)

Input Parameters

key name	type	name	description
result_name	str	Result name	Name of the result dataset.
seed *	int	Seed	Initial value to seed the random number generator.

Input Datasets

key name	type	name	description
building_dmg *	ergo:buildingDamageVer4 ergo:buildingDamageVer5	Building damage	A building damage dataset.
housing_unit_allocation *	incore:housingUnitAllocation	Housing allocation	A housing unit allocation dataset.
block_group_data *	incore:blockGroupData	Block group data	A block group racial distribution dataset.
value_loss_param *	incore:valueLossParam	Loss parameters	A table with value loss beta distribution parameters.

Output Datasets

key name	type	name	description
result *	incore:popDislocation	Results	A dataset containing results (format: CSV).

(* required)

Execution

code snippet:

    # Create population dislocatin
    pop_dis = PopulationDislocation(client)

    # Load input dataset
    pop_dis.load_remote_input_dataset("building_dmg", building_dmg)
    pop_dis.load_remote_input_dataset("housing_unit_allocation", housing_unit_alloc)
    pop_dis.load_remote_input_dataset("block_group_data", bg_data)
    pop_dis.load_remote_input_dataset("value_loss_param", value_loss)

    # Specify the result name
    result_name = "IN-CORE_1bv6_population_dislocation"

    seed = 1111

    # Set analysis parameters
    pop_dis.set_parameter("result_name", result_name)
    pop_dis.set_parameter("seed", seed)

    # Run Population dislocation analysis
    pop_dis.run_analysis()

Jupyter notebook: populationdislocation.ipynb