-
Notifications
You must be signed in to change notification settings - Fork 5
NED
NED (the New Economic Demographics module) is the major step of SWIM, after SI which simply preps input files to be used in SWIM. NED is the second generation of SWIM's economic module. NED follows ED, which was just the Economic Demographics module. The NED module as used in SWIM is fairly simplistic. Five, mostly static, csv files are provided to NED. Those files have model-wide (not by zone, but by the entire model region) economic and demographic totals for each year of SWIM. Those totals into the future are held constant unless NED feedback is turned on, in which case, changes in AA utilities between the current scenario and a reference scenario will cause the model-wide economic totals to change through time based on the differences between scenarios in their utilities. If feedback is not on, NED is as simple as providing the values from the csv files that have been input to SWIM to the given SWIM year to start off the land use and other modules in SWIM. So from a SWIM operational standpoint NED is very straight forward. However, that simplicity is misleading. The real complexity of NED is in developing those model-wide economic and demographic totals through time by hundreds of sectors. The following text explains that process for preparing the files for NED and how those files are used in SWIM.
The NED module is a Python program located in scenario_name/model/code/NED.py. The ModelEntry NED instance calls the com.pb.tlumip.ed.NEDModel class, which initiates the Python program as a process, redirects its console output to the appropriate loggers, and indicates if the module finished normally or not. The specifics of how the process commands are formed come from the following property file keys:
-
python.executable- the path to the Python interpreter -
ned.python.command- the path to the NED.py program file -
ned.property- the path to the NED-specific property file
The New Economics and Demographics (NED) Module replaces the original Economics and Demographics (ED) module. The ED module produced forecasts that were internally consistent, but not necessarily consistent with forecasts from any other model. As a result, a lot of effort went into overriding ED forecasts to develop reference scenarios that were consistent with official state forecasts, which changed every quarter.
The NED module is organized around the concept of scenarios. A scenario is a complete set of NED output that is intended to represent the economic and demographic aspects of a particular future to be modeled. NED outputs are all at the model-region-wide level and include forecasted output and employment by AA activity, imports and exports by AA commodity, and population by five-year age categories. Construction activity and government revenue forecasts are reported separately as well. All dollar values are expressed in 2009 base-year dollars, consistent with the rest of SWIM2.
The NED module has two main components:
-
An exogenously-defined reference scenario that is consistent with assumptions that drive ongoing planning efforts in Oregon.
-
A feedback mechanism that allows transportation and land-use policy changes or other shocks (e.g., natural disasters) to drive deviations from the way levels of economic activity change over time under the assumptions that underlie the reference scenario.
The NED module does not include an economic or demographic forecasting model of its own. It consists instead of a representation of the output from exogenous economic and demographic models and the ability to modify those values in appropriate and consistent ways in response to deviations from the reference case in the scenario being evaluated.
The economic feedback mechanism allows differences between prior-year AA model-wide composite utilities from the current scenario and the reference scenario to influence NED forecasts, prior to their use by other modules. The feedback module applies a user-defined, assumed elasticity between percent change in region-wide location utility and percent change in economic activity levels. If this elasticity is set to zero, feedback does not happen and NED produces output identical to the reference scenario. Feedback also does not happen in historical model years (those before the base year data used to construct the reference scenario).
NED operates exclusively at the model-wide level. NED produces estimates of employment and output by AA activity, of imports and exports by AA commodity*, of residential and non-residential construction activity, of government revenue in three categories (federal, state and local, and corporate), and of population by five-year age groups.
'*' = "Output" is a standard term in economics meaning the dollar value of what is produced. For example, the output of the food manufacturing industry in a region is the dollar value of food manufactured in that region. AA uses the term, "activity" to mean "economic sector" or "industry". Units for output in each "activity" are measured in dollars. Units of employment in each "activity" are measured in FTEs.
The software implementation of the NED module is a Python script. Each time it is run, it is given a pointer to a properties file, which it reads to obtain the following parameters:
| Property | Description |
|---|---|
| ned.input.directory | where reference scenario files are read from |
| ned.activity forecast.path | where model-year activity forecast is written |
| ned.trade forecast.path | where model-year trade forecast is written |
| ned.construction forecast.path | where model-year construction forecast is written |
| ned.population forecast.path | where model-year population forecast is written |
| ned.government forecast.path | where model-year government forecast is written |
| ned.prior activity forecast.path | where prior-year activity forecast is read from |
| ned.prior trade forecast.path | where prior-year trade forecast is read from |
| ned.prior construction forecast.path | where prior-year construction forecast is read from |
| ned.prior population forecast.path | where prior-year population forecast is read from |
| ned.prior government forecast.path | where prior-year government forecast is read from |
| ned.base.year | e.g., 2009 |
| ned.model.year | an integer value; not a calendar year |
| ned.base.year.model.year | the model year identifier for the base year, e.g., 19 |
| ned.prior_activity_summary.path | where the prior year's composite utilities from the current scenario are read from |
| ned.reference_prior_activity_summary.path | where the prior year's composite utilities from the reference scenario are read from |
| ned.feedback_elasticity | zero for no feedback, 1 for a 1-to-1 change in regional composite utilities to economic output |
| ned.reference_scenario_base_year | e.g., 2016; the most-recent year with IMPLAN data |
Files that NED generates for other modules are written to the appropriate directories as specified in the ned.*_forecast.path parameters. However, the baseline forecast data that NED uses over time resides in a single directory, specified by the ned.input.directory parameter. The files in that directory are in the same format as the output files except that they have an extra column on the left, in which the year is specified. All NED inputs and outputs represent the entire model region (including Oregon and the halo) and all dollar values are in base-year dollars.
ned.feedback_elasticity = 0 will result in no NED feedback. A value of greater than 0 will turn on NED feedback and the feedback will be active in the scenario years that are later than ned.ref_scenario_base_year. For example, if ned.ref_scenario_base_year=2016 then a non-zero elasticity value would make feedback active in t27 (2017) and later years.
In the base year, the NED module reads values* for that year from the Baseline Scenario files and writes them to that year's NED model files. In subsequent model years, it reads that year's and the prior year's baseline forecast from the Baseline Scenario files, the prior year's model-run forecast, and the prior-year values of feedback variables (composite utility and size from the current run and from a run of the reference scenario) from files written by AA. (* =The Reference Scenario consists of activity/commodity names, years, and values. These values are all numbers. For example, the output of the MFG_food_hi activity in 2020 is 11,275,526,635 in the reference scenario. So 11,275,526,635 is the "value" in that instance.)
For each variable in each forecast, the NED module calculates the ratio of this year's value to last year's in the reference scenario. If feedback is enabled, it then adjusts those ratios using the fed-back values from AA combined with an assumed elasticity). The elasticity describes the relationship between percent difference in the prior year's region-wide composite utility for an activity between the current scenario and the reference scenario and the proportional difference in level of that activity in the current year in the current scenario compared to the reference scenario. If the assumed elasticity is zero, feedback has no effect.
The percent change in utility is measured by comparing composite utilities for the prior model year from the current scenario to those from the reference scenario for each activity. The relative change in utility ((current scenario's composite utility minus reference scenario's composite utility) divided by the reference scenario's composite utility) is multiplied by the elasticity and then added to one to produce a factor by which the ratio of this year's activity level to the prior year's is scaled before applying the resulting adjusted ratio to last year's activity in the current scenario to get this year's activity in the current scenario. The model-year forecasts are then written to the appropriate directories.
For example, if in the current scenario, prior-year employment in a sector were 100,000 and employment in the reference scenario grew by 3.0% from the prior year, without feedback the current-year employment would be 103,000. If feedback were enabled with an elasticity of 1.05, and the prior-year composite utility for that sector in the current scenario were 2% lower than in the reference scenario, then current-year employment would be adjusted by -2.1% (1.05 times -2%), reducing it to 100,837, or an overall change of +0.837% ((1 + 3%) * (1 - 2.1%) - 1).
The effects of feedback compound over time. If, for example, composite utilities in the current scenario increase from those in the reference scenario and remain higher, activity levels in the current scenario will rise above those in the reference scenario, and the difference between scenarios will become larger each year as the effects compound from year to year.
The feedback elasticity measures how responsive NED's representation of the economy of the entire model area is to changing attractiveness for the various activities as represented by regional composite utilities for each activity coming from AA in the prior model year. It is assumed that each activity is equally responsive, so only one elasticity is specified by the user (set as property ned.feedback_elasticity; see NEW properties section above). When first establishing or adjusting this elasticity, it is advised to start at 1.0. An elasticity of 1.0 implies, for example, that a 10% increase in attractiveness for a particular activity will result in a 10% increase in the level of that activity in the following year.
If 1.0 is found to make NED's representation of the regional economy too responsive or not responsive enough, revise the assumption by increments of 0.05. If NED's representation of the economy is too responsive to changes in utilities, start reducing the assumed elasticity from 1.0, to 0.95, then 0.90, then 0.85 and so on, until the desired level of responsiveness is reached. If NED's representation of the economy is not responsive enough to changes in composite utilities, revise the elasticity up to 1.05, then 1.10 and so on, until the desired level of responsiveness is reached.
As of Dec 2020, the initial testing approach described above was tested. The test was completed on the most extreme scenario under the "Rough Roads Ahead" (RRA) work conducted by ODOT. In that work, prior to NED feedback being a feature in SWIM, SWIM was used to estimate an economic impact of roughly 100,000 jobs in Oregon and a loss of state gross product of roughly $200B over 20 years of highway disinvestment. Five feedback scenarios were tested; 0.5, 0.75, 1.0, 1.2, and 7.0. The 7.0 was an early extreme test, and it caused the model to crash within a few years (due to extreme economic shifting/changes). So 7.0 was discard from the overall review and comparison. Similarly, the factor of 1.2 also created run issues, but not until further out years (past 2040), so a factor of 1.2 might work in some cases, but it should be known that in more extreme scenarios 1.2 might not run all the way through time.
The table below shows the regional and statewide impact to loss of jobs after 20 years under the RRA scenario under 4 feedback settings. The table shows that the feedback setting of 0.5 was the closest to past work and published findings.
| Feedback 1.2 | Feedback 1.0 | Feedback 0.75 | Feedback 0.5 | |
|---|---|---|---|---|
| CoastTotal | -54,114 | -52,151 | -50,814 | -48,764 |
| Metro | -57,932 | -45,464 | -29,593 | -9,947 |
| MWV | -46,252 | -36,462 | -24,783 | -14,985 |
| SV | -24,789 | -19,649 | -13,045 | -6,948 |
| Central | -19,301 | -14,924 | -9,808 | -4,857 |
| Eastern | -20,576 | -18,532 | -15,601 | -13,518 |
| Oregon Total | -222,964 | -187,182 | -143,644 | -99,019 |
However, when review the loss to statewide GDP over the 20 years choosing 0.5 is not as clear. Under the feedback setting of 0.5 The state see a loss of $120B, but both the Metro and the Central Oregon region see positive growth, even though these areas have disinvestment in the network. This is likely because Metro and Central Oregon have relatively mild disinvestment when compared to the rest of the state, and while they are impacted the model has the potential to see these areas as relatively pretty well off when compared to other regions (zones) in the state). As the feedback setting is increased to 0.75 and 1.0 the overall impact and the regional impact is better aligned with the expectations for how RRA would impact the state (all regions would see a decrease, some worse than others).
| Feedback 1.2 | Feedback 1.0 | Feedback 0.75 | Feedback 0.5 | |
|---|---|---|---|---|
| CoastTotal | $-112B | $-110B | $-108B | $-106B |
| Metro | $-81B | $-53B | $-22B | $6B |
| MWV | $-54B | $-39B | $-21B | $-5B |
| SV | $-25B | $-18B | $-9B | $-1B |
| Central | $-18B | $-11B | $-5B | $2B |
| Eastern | $-27B | $-24B | $-19B | $-16B |
| Oregon Total | $-317B | $-255B | $-184B | $-120B |
Based on these runs for RRA a setting of 0.5 seems a little too light, 1.0 seems a little too strong, and 0.75 seems just right. However, this is just initial guidance on setting feedback as of December 2020. The team plans to continue to test how feedback performs, under additional scenarios and considering and evaluating a wider array of model outputs. Planned scenarios at this time are to review feedback under different land use scenarios as well as different pricing and tolling scenarios. As that work is completed, this wiki and feedback guidance in general is expected to be improved. As of this date the guidance is to either set feedback at 0.75 for scenarios where feedback is desired to be used. Or, better yet, to run feedback scenarios at both 0.5 and 1.0 and report the range of impacts. Providing a range is the recommended best practice moving into the future.
The reference scenario is built from external data and forecasts. It starts with IMPLAN data for the model region for base year output, employment, imports, exports, and population. IMPLAN populations are subdivided into five-year age groups using Census data. The IMPLAN model region used includes exactly the same counties as the SWIM2 model region (all 36 Oregon Counties plus halo counties).
The software implementation of the reference scenario generator uses the official employment forecast for Oregon and a national forecast that substitutes for the state forecast in years for which the state forecast is not available. Oregon provides an official forecast that goes out at least eight years. It is assumed that the actual growth rates in the halo region will more closely resemble Oregon's than the rest of the nation, so Oregon's growth rates are applied to the entire model region in the years they are available. The national forecast is produced by IHS Markit (formerly known as Global Insight) and purchased by the State of Oregon to drive its own forecasting models, including the official state forecast that is used for Oregon. This is done to help ensure that the Oregon forecast and the national forecast will be consistent with each other.
Employment is forecast by IMPLAN sector (536 sectors) using a crosswalk that matches one exogenous forecast sector to each IMPLAN sector. The exogenous forecasts from the Oregon Department of Administrative services Office of Economic Analysis and IHS Markit have many fewer sectors than IMPLAN and each has its own set of sectors and its own crosswalk. Each year's employment in each IMPLAN sector is forecasted by applying the ratio of that year's employment to the prior year's employment in the corresponding sector in the exogenous forecast to the prior year's employment in that IMPLAN sector.
Employment by IMPLAN sector is aggregated to employment by AA activity (52 combinations of economic sector and space type) using a crosswalk that allows IMPLAN sectors to be split among AA activities, if necessary.
For each sector in the national forecast, the ratio of labor productivity (output per employee) in the current year to labor productivity in the prior year is calculated. For each IMPLAN sector in the employment forecast, the minimum of this ratio or 1.05 is applied to the prior year's labor productivity within the model region and the resulting, updated labor productivity is multiplied by the forecast of employment for the current year in the model region, yielding forecasted output for that IMPLAN sector in the model region in the current year. Output by IMPLAN sector is then aggregated to output by AA activity (52 activities) using a crosswalk that allows IMPLAN sectors to be split among AA activities (for the few IMPLAN sectors that require a split). The crosswalks for output and employment differ from each other as industry employment may be spread across space types differently than output. For example, for a manufacturing industry, part of employment may be assigned to office space whereas all output is assigned to factory space.
For each IMPLAN sector, the ratio of the current year's output to the base year's output is calculated. Each industry's make of each export commodity from the base-year IMPLAN structural matrices is then scaled by that sector's output ratio to estimate exports by that industry in the current year. Each industry's use of each import commodity from the base-year IMPLAN structural matrices is scaled by that sector's ratio to estimate imports by that industry.
The ratio of current-year population to base-year population is calculated and applied to institutional imports and exports by commodity from the base-year IMPLAN structural matrices to estimate current year imports and exports by institutions.
Industry and institution imports and exports are aggregated by IMPLAN commodity (536 commodities) and then aggregated to AA commodities (54 commodities) using a crosswalk that allows IMPLAN commodities to be split among AA commodities if necessary.
Construction forecasts are a re-arrangement of the construction output dollars into those associated with residential structures and non-residential structures. The re-arrangement is completed as a preparatory step for the ALD module. The output of other non-residential construction activity, such as road building, is excluded.
Government forecasts use elasticities of government revenue with respect to total employment to estimate state and local government revenue, federal revenues from corporate taxes, and federal revenues from personal income taxes.
The NED Reference Scenario Generator runs outside of the SWIM2 modeling framework and produces the NED input files that must exist before the SWIM2 model is run. It gathers data from several IMPLAN output files, from a state economic forecast file, from the IHS Markit national long-run economic forecast file, and from files containing crosswalks and sector mappings. The IMPLAN and state forecast files are in their original format, so the Baseline Scenario can be updated by substituting newer copies of the forecast files and then rerunning the Baseline Scenario Generator, without any need to modify the new forecast files. Meaningful changes in the format of the forecast files by the entity that produces them will require either modifying the code or reformatting the input file.
The base year for the reference scenario generator may be different from that of the SWIM2 model. It is the most recent year for which all input data are available. The reference scenario generator will backcast from its base year to produce reference-scenario data for all years needed by SWIM2. It also can forecast beyond the end of the national forecast by assuming that the economy will continue to grow at the same rate as in the last year of that forecast.
The reference scenario generator starts by building internal data structures to hold input values and the results of intermediate calculations. It then reads from various input files and puts their data into the internal data structures.
To prepare the input files produced by IMPLAN, the analyst must first obtain IMPLAN data sets for the state of Oregon and all counties in the halo. The analyst will then use the stand-alone version of IMPLAN (not the online tool) to build an IMPLAN region named "Oregon and Halo", which includes all those counties and to build a Social Accounting Matrix for that region. Once the social accounting matrix is built, the analyst will produce the "Industry Detail" file for that region, which will be named "Oregon and Halo Industry Detail.csv" by IMPLAN.
Activate the "Explore" menu in IMPLAN by selecting "Advanced Modeling" in User Preference (Analysis tab). To extract the industry detail file, click on "Study Area Data" under the "Explore" section, then click on "Export".
The reference scenario generator also requires additional IMPLAN files that describe imports and exports (from/to the rest of the US and the rest of the world) of commodities by industries and institutions. These files are part of a set of files IMPLAN can produce for computable general equilibrium (CGE) models.
To extract the 26 CGE files:
- Click on "Social Accounts" under the "Explore" section.
- Click the tab that reads "IxC Social Accounting Matrix"
- Click "Export"
- Click "Industry detail SAM files (CSV only)"
- Click "GAMS 26 File"
The reference scenario generator needs eight of those 26 tab-delimited files:
- 1x7.dat
- 1x8.dat
- 4x7.dat
- 4x8.dat
- 7x1.dat
- 7x4.dat
- 8x1.dat
- 8x4.dat
All required input files need to be moved to the same file directory as make_base_scenario.py before it runs.
The Excel workbook "employment-annual.xls" may be downloaded from the Oregon Office of Economic Analysis web page. Verify that the rows and columns are the same as in the included sample. If they aren't, either add or delete blank rows or columns as needed, or edit make_base_scenario.py as necessary. The layout of this file rarely changes.
The Excel workbook "County_forecast_March_2013.xls" may also be downloaded from the Oregon Office of Economic Analysis web page. This file has remained unchanged since 2013 and responsibility for producing it has shifted to Portland State University.
For halo counties, the 2010 populations may be obtained from American Fact Finder web page at census.gov.
The file "pop_forecast.xlsx" is also required. This file was prepared from data in "County_forecast_March_2013.xls", along with 2010 census populations for halo counties. It is an input to the reference scenario generator, and needs to be in the same file directory as make_base_scenario.py, but does not need to be altered until a new long-range population forecast for Oregon becomes available.
A file similar to "2018 Feb Long Run Baseline.xlsx" may be obtained through ODOT's Transportation Economist. This file is very large. Two of the 56 tabs contain data needed for the reference scenario generator. As of 2018, those are named Emp1A and DataA. The file "IHS Markit Forecasts.xlsx" contains the relevant portions of these tabs and is an input to the reference scenario generator, and needs to be in the same file directory as make_base_scenario.py. Copy all rows beginning with "Total Nonfarm Payrolls" from Emp1A into the Employment tab of "IHS Markit Forecasts.xlsx" and then delete any blank rows. From DataA, copy all rows beginning with "Generated Output" (rows 446 to 517 in February 2018) and paste into row 2 of the Output tab in "IHS Markit Forecasts.xlsx".
These files are included with the reference scenario generator and need only be placed unaltered into the same file directory as make_base_scenario.py.
- implan_aa_commodity.csv maps 536 IMPLAN commodities to 54 AA commodities
- implan_aa_employment.csv maps 536 IMPLAN industries to 52 AA activities
- implan_aa_output.csv maps 536 IMPLAN industries to 52 AA activities
- implan_gi_employment.csv maps 536 IMPLAN industries to 46 IHS Markit employment sectors
- implan_gi_output.csv maps 536 IMPLAN industries to 68 IHS Markit industries
- implan_oea_employment.csv maps 536 IMPLAN industries to 22 OEA employment sectors
Most of these won't need to be changed often.
- BASE_YEAR = 2016 The year of the IMPLAN data used; change as needed
- BEGIN_YEAR = 2009 The base year for SWIM2
- END_YEAR = 2050 The last year to be forecasted
- LAST_FORECAST_YEAR = 2048 The last year in the national forecast used
- MAX_PRODUCTIVITY_RATIO = 1.05 (Global Insight data sometimes implies that output per employee can save an annual growth of greater than 5%. That was found produce an unreasonable level of output for those sectors. 5% (or 1.05) is a limiting factor on larger (less likely) productivity increases).
- BASE_FED_TAX = 93817953918.0 From HBA
- BASE_SL_TAX = 71879103322.0 From HBA
- BASE_CORP_TAX = 69549704567.0 From HBA
- FED_TAX_ELASTICITY = 1.214 From HBA
- SL_TAX_ELASTICITY = 1.234 From HBA
- CORP_TAX_ELASTICITY = 1.054 From HBA
- BASE_YEAR_GDP_DEFLATOR = 105.899 From Bureau of Economic Analysis; change if BASE_YEAR changes
- BEGIN_YEAR_GDP_DEFLATOR = 94.999 From Bureau of Economic Analysis; change if BEGIN_YEAR changes
Note that the two GDP deflators must have the same base year (e.g., 2012 as is the base year for currently used/published by Bureau of Economic Analysis, this base year will change periodically) as each other.
Base year results for the Baseline Scenario are calculated by aggregating IMPLAN data to the categories used in SWIM2. IMPLAN population for the model region is attributed to five-year age groups based on 2010 census data proportions.
Employment in each subsequent year is then forecasted by applying the appropriate growth rate from the State's forecast to the prior year's employment. If the model year is beyond the end of the State's forecast, the appropriate growth rate from the national forecast is used. Employment is then aggregated over IMPLAN sectors to SWIM2 categories.
The State does not forecast output. To forecast output, the ratios of output per employee in each sector in the current year to that in the prior year from the national model is calculated and applied to the prior year's output per employee for the region. The adjusted output per employee is then multiplied by forecasted employment to obtain a forecast of output.
Population is forecasted by applying growth rates by five-year age group from the Oregon population forecast to the prior year's model-region population. Population forecasts specific to the halo portions of neighboring states have not been obtained or developed, and it is therefore assumed that population growth rates in those counties would be more similar to those in Oregon than to the rest of their states or the nation.
For the trade forecast, industry imports and exports are forecasted separately from institution imports and exports. Industry exports are forecasted by taking the ratio of output for each industry in the model year and the base year and applying it to the exports of each commodity made by that industry in the base year. Industry imports are forecasted by taking the ratio of output for each industry in the model year to that in the base year and applying it to the imports of each commodity used by that industry in the base year. Imports and exports by institutions (e.g., households and governments) are calculated similarly, but using population ratios rather than output ratios.
The construction forecast used by ALD is constructed by aggregating output dollars from IMPLAN construction sectors, leaving out construction of things other than houses and buildings (e.g., construction of roads is not included).
The government forecast uses estimated elasticities relating relative change in government revenues to relative change in employment, which are applied to the change in forecasted employment to get change in government revenue, which is then applied to the prior year's government revenue. The estimated elasticities used are stored in named constants near the top of the script.
SWIM2 currently expects dollar values to be in BEGIN_YEAR (2009) dollars. At this point, all dollar values in the reference scenario generator's tables are in BASE_YEAR (2016) dollars. As a part of the writing of output files, dollar values are converted to BEGIN_YEAR dollars using the ratio of BEGIN_YEAR_GDP_DEFLATOR to BASE_YEAR_GDP_DEFLATOR (0.8971).
[Developed by HBA, have John provided some additional instruction and context]
NED as currently implemented is primarily an exogenous forecast. As such, the Reference Scenario has no estimated parameters, and makes just a few assumptions noted below.
The estimated elasticities used in generating the government forecast in the reference scenario are:
- FED TAX ELASTICITY = 1.214
- SL TAX ELASTICITY = 1.234
- CORP TAX ELASTICITY = 1.054
These elasticities relate relative changes in forecasted employment to relative changes in forecasted government tax revenues.
| Data Element | File(s) | Source |
|---|---|---|
| Reference scenario Forecasts for all years | activity_forecast.csv | Exogenous |
| " | trade_forecast.csv | " |
| " | government_forecast.csv | " |
| " | construction_forecast.csv | " |
| " | population_forecast.csv | " |
| Modelwide composite utilities of production activity in reference scenario | reference_activity_summary.csv | AA from prior run of reference scenario |
| Modelwide composite utilities of production activity in current scenario | activity_summary.csv | AA from current run |
Note: composite utilities are required only when feedback will be used.
SWIM-TLUMIP Model User Guide, version 2.5
- SI - SWIM Inputs
- NED - New Economic Demographics
- ALD - Aggregate Land Development
- AA - Activity Allocation
- POPSIMSPG - PopulationSim Synthetic Population Generator
- PT - Person Transport
- CT - Commercial Transport
- TA - Traffic Assignment
- TR - Transit Assignment
- SL - Select Link
- SWIM VIZ - Reporting DB