This project is an application of the GTAP-InVEST earth economy modeling approach to estimate global-local-global estimates of the influence of ecosystem services on a computable general equilibrium of a national economy. Below are the main steps performed in this analysis as well as a discussion of the results.
To change the input definitions for the SEALS standard run, I need to modify the scenario_definitions.csv file called on line 62 of the run_test_standard.py: p.scenario_definitions_path = os.path.join(p.input_dir, 'scenario_defininitions.csv'). Under aoi, I changed the country alpha-3 code to BLZ for Belize. The scenerios are already set to baseline and SSP2 under exogenous_label. Here I duplicate a line from SSP2 to create SSP1 and SSP5 while removing SSP2. To modify the years that SEALS generates the LULC maps, I insert 2030 2035 2040 under years for the SSP1 and SSP5 scenerios while keeping the baseline year at 2017 (this is the baseline year of data that SEALS is pulling from -- so don't change). Note that the labels in the scenerio definitions are important. So be sure to relabel them consistently as shown in the scenario_definitions.csv file in this repo.
Finally, to run the new SSP scenerios you will need to download the correct coarse_projections_input_path from Land Use Harmonization data. You will need to approriately save this in the base_data using a consistent directory structure from the baseline input data.
For hazelbean to run parallel processing, it creates separate project folders given the name given in line 24. If you don't change the name, run_test_standard.py will throw an error. So, instead create a new file for the Belize run called run_BLZ_standard.py and change project_name = 'BLZ_standard'
Once completed I now I have a new set of scenerio definitions which I an use in the run_BLZ_standard.py. You can examine this file as scenario_defininitions1a.csv. To examine the results, you can now look in the seals folder under projects and you should find the LULC maps (and other output) in a folder called BLZ_standard alongside the initial folder for the initial test run called test_standard.
You will want to put the scenario_definitions.csv file into a new project folder labeled BLZ_standard\input before running the run_BLZ_standard.py. If you do not, then it will populate the standard test run with Rwanda instead of the new country (Belize). When you go to run this script, run it within the Debug function with VSCode using the internalConsole. Make sure to working within the earth_economy_devstack (Workspace) and to run with your teem_devstack python kernel so you have all the correct modules.
I implement a policy that supposes that there are no conservation requirements. This is to say that protected areas can be converted to any land use type. To apply this policy, I edit the strict_pa attribute of the default_global_coefficients.csv from the seals base data. This change can be found in the attached default_global_coefficients.csv file. Below are the LULC as a result of this change.
Provide a narrative of how the different scenerios changed when this land use class was expanded or contracted.
The SSP1 scenario had far more change than the SSP5 scenario; which was suprising. The main finding was that if we stay in a world of conservation, nature will expand in several areas of Belize. This is to suggest that urbanization and crop expansion in Belize has a minimal likilihood to increase under any SSP scenerio. But, Belize is an important area where natural captial could expand if we maintain a more enviornmental SSP.
Using the first set of LULC maps from SEALS, assess ecosystem service provision for carbon storage, water yield, pollination, sediment retention, and nutrient retention using the InVEST ecosystem service modeling.
I use the LULC maps generated through SEALS as inputs to assess ecosystem service provision through the InVEST model. Before using the generated LULC maps in the InVEST to compute the ecosystem services, we need to make sure that the projection systems used in the maps are confirmable. The SEAL output uses WGS84 and INVEST uses EPSG:5235. Another commonly used and preferred projection system is Robinson projection (ESRI:54030). Additionally, to run the five InVEST models you need to collect the global input data required for each process based model from Chaplin-Kramer et al. "Global modeling of nature’s contributions to people".
The input base data for the InVEST models can be found in the base-data folder. To create the following images, you can run the run_invest_BLZ.py script.
There is no visual distinction between the different scenerios. This exercise shows that Belize will remain unchanged in the near term for various SSP scenerios. It could also be that 2030 -- 2045 is too short of a timeframe to see meaniful change to the landuse (and therefore the ecosystem service provision) of the country.
Write an executive summary for what the policy maker interested in "green economic development" should know.
In short, Belize's ecosystem servies will be unaffected by various climate change scenerios in the near term through 2045. The country of Belize has natural capital which will not vary much or contribute much likely because it does not have land use that can be changed by the small population. Policy makers interested in 'green economic development' do not a lot of options for expanding ecosystem services as the land is not used much for cropland or urban spaces. And even in a scenerio of industrial expansion, there is little comparative advantage to expand industry in Belize. So, there is little concern with what can be done with the current LULC of the country. I would instead say these policy makers can focus on other issues.
This README file so far has described the final project analysis and the results. In addition, for the APEC 8601 natural resource course, we produced some empirical results using python code. I elaborate on this below.
- In problem set 02 I work with the DICE model for optimal production considering the damages from global warming. Additionally, I work with the MAGICC model to show the projected global temperature under different RCP scenerios through 2100.
- In problem set 04 I run the InVEST ecosystem services model to estimate the ecosystem service benefits from crop pollination.
This repository has been made public and can be found here.