ARIES for PEOPLE Explorer users guide

ARIES for PEOPLE Explorer

The ARIES for PEOPLE Explorer is a web-based application built on the k.LAB Integrated Modelling Platform. The application has access to all information (data and models) available on the Integrated Modelling network, and provides a dedicated user interface to allow to the pilot countries to easily access and test the output of the PEOPLE-EA project, funded by the European Space Agency, and developed by the ARIES team (BC3) in collaboration with the VITO team and using the OpenEO (Open Earth Observations) platform.

I. Spatial and temporal context of the analysis

At the top of the menu on the left side of the screen, you can specify the geographic area and temporal and spatial scale.

1. Where

   At the top of the panel, there a drop-down menu provides three options to select an analysis context by zooming and panning on the map. When the "administrative regions" or "river basin" option is chosen, the currently highlighted context will be outlined in light blue.
   1. Map boundaries: Select an area of interest by panning and zooming in/out. The entire area visible on the screen becomes your analysis context.
   2. Administrative regions: This option automatically identifies the largest administrative entity (e.g., Country or Subnational Unit) in the area selected, according to the M49 standard endorsed by the UN. By zooming in, the user can choose a smaller administrative region. We recommend this option for novice users, as it offers a simple way to identify standard administrative boundaries for ecosystem accounting.
   3. River basin: This option selects an area of land draining to a specific water body based on FAO Hydrological Basins (simplified).
      
      
      
      
   4. Alternatively, the user can also directly type the name of a geographical context (i.e., country, region, city, or other geographic entity) in the ARIES for PEOPLE Search bar, starting with a capital letter. 
      These names are queried from the OpenStreetMap (OSM) database. Users should be aware that OSM boundaries may differ slightly from those selected using the "Administrative regions" option of the drop-down menu (i.e. a context selected with this option include the country's territorial waters)
       
   5. Countries are warmly invited to provide their official boundaries in case the default options currently integrated in ARIES to identify geographical boundaries don not meet the expected NSO's specification.

2. Resolution

   The user can select the spatial resolution for analysis, and choose between meters or kilometers.
   
   If a resolution is chosen that is higher than that of available input data, ARIES will compile accounts at the resolution of the finest grained available data.
   
   
   

3. When

   Select the year(s) for your analysis. If data are missing for a specific year of interest, ARIES automatically fills gaps using the closest available year's data 

   Still holds true?

   (To be confirmed given the behaviour of OpenEO UDP)

   
   
   
   I. Single-year analysis (uncheck the box)
   
   
   II. Multi-year analysis - showing change over time (check the box)
   
   
   
   

4. Reset

   The red "X" button on the upper right can be used to reset a previously selected context, or to stop a computation in progress (all computed results will be lost).
   

II. Study setup

The People EA deliverables section contains the list of the variables, in the current state of development of the project, that can be observed by simply clicking on the variable's box.

This action will trigger the model(s) underlying the variable(s) selected, in case more variables are selected, the computational flow is continuous and after the whole workflow to compute and a variable is completed, the system moves to the next tasks in the queue of computation. The models to be computed follow the order of selection of the user.

In case to compute a variable an input used in a previous computation is required, this is not re-calculated, so you should expect the additional variables computed to be quicker, in case they share inputs with previously computed variable. Nevertheless, most of the variable are associated to a different dataset, so for this specific case the time of computation should be similar. 

Forest Ecosystem Condition accounts

There are 3 main outputs of the condition accounts:

• Variable (descriptor) = the original ecological measurement used to value a certain aspect of condition 
• Indicator = the rescaled ecological measurement used to value and compare a particular dimension of condition captured in the variable.The rescaling occurrs comparing the value from the original measurement (variable) against the same measurement (same variable) in forests with similar characteristics.
• Index = the weighted value of an indicator in the overall condition assessment

Forest ecosystem condition variables

Forest ecosystem condition indicators

Each variable is scaled between 0 and 1, with values getting closer to 1 as the forest conditions improve.
This allows each indicator to represent in a consistent way the information coming from the variables, which in their original values have different ranges (i.e. assuming values from -100 to 750) but also different interpretation of such values, and not always higher value are associated to better conditions (e.g. if we look at variable like Drought Severity or Forest Fragmentation, higher values are associated to worse forests conditions). 

There are 7 indicators available in the application:

1. A1.1 Normalized Difference Water Index

2. A2.1 Soil Organic Carbon

3. B1.1 Threatened Forest Bird Species diversity

4. B2.1 Leaf Area Index

5. B3.1 Net Primary Production

6. B3.6 Burned severity

7. C1.2 Forest Connectivity

Reference areas

These areas are used to identify forest considered in good health, or in bad shape, so as to define the upper and lower bounds of the variable for the reference year, which are used to rescale variables into indicators. The years of reference is the first available in the Corine dataset, the year 2000.

Input data

• vector - primary forests & protected areas
  • The EPFD v2.0 - https://www.nature.com/articles/s41597-021-00988-7
  • IUCN data - https://www.protectedplanet.net/en/thematic-areas/wdpa?tab=WDPA
• vector - corine year 2000
• raster - corine acc year 2000
• raster - HRF forest 100m year 2012
• vector - geophysical regions EU
• raster & vector - EU countries & admin bounds

Forest ecosystem types

European forests have been categorized based on the landcover forest class, as per the Corine datasets, which distinguish 4 types of forest: 

1. Broad-leave forest (311)
2. Coniferous forest (312)
3. Mixed forest (313)
4. Transitional woodland & shrub (324)

and combined with the bioregion information, w

• create 44 forest ecosystem types following the definition of Maes et al. (2023) Maes, J., et al. (2023). "Accounting for forest condition in Europe based on an international statistical standard." Nature Communications 14(1).
• load the corine vector layer in QGIS (Version 2020 20u1 (dated 02/2020): First Update of Version 20)
  1. select only the corine forest classes
     1. broad-leave forest (311)
     2. coniferous forest (312)
     3. mixed forest (313)
     4. Transitional woodland & shrub (324)
  2. save to disk as corine_forest_2000.gpkg
• load the eu biographic zone file (https://sdi.eea.europa.eu/data/def7ac06-7d3f-4da5-880c-a76a73953cfc)
  • create the alpine Scandinavian zone in vector file - just split the alpine multi-polygon into single ones → mark the Scandinavian areas and dissolve into an own zone
  • 11 biogeographic zones are of interest
    1. Alpine,
    2. Arctic,
    3. Atlantic,
    4. Black Sea,
    5. Boreal,
    6. Continental,
    7. Macaronesian,
    8. Mediterranean,
    9. Pannonian,
    10. Steppic regions
    11. Alpine (Scandinavia) → forest in scandinavian mountains on border between Norway and Sweden)
  • save to disk as biogeographic_zones_maes-article.gpkg
• intersect the corine and biogeographic zone files → forest ecosystem types file save as forest_ecosystem_types_maes-article_v2.gpkg

• run a raster process to generate the final forest ecosystem map → attributing by the geographic zone id using the LUT for geographic zone [raster value, geographic zone name] and  LUT for corine [raster value, corine class])

• final raster value to forest eco has to be written to LUT

• • • Extent Accounts

      These accounts measure the extent of the IUCN Global Ecosystem Typology ecosystem types, or land cover, present in the context of your analysis, in km².
      The different types of accounts provide varying levels of detail in summarizing ecosystem/land cover extent and its change over the selected time period.
      
      
      
      

    • Condition Accounts

      These accounts measure ecosystem condition. Currently, only forest ecosystem condition accounts are supported, but condition accounts for other ecosystem types will be added soon (beginning with those for grasslands).
        
      
      
      The conditions metrics available for inclusion in the account appear in a drop-down menu when the user clicks on the triangle next to "Forest condition metrics".
      
      
      
      Three types of ecosystem condition accounts are available:
      
      1. Condition Variable Account: report the value of each condition metric in their originally observed values (non-transformed);
      2. Condition Indicator Account: rescale ecosystem condition variables to values between 0 and 1. Rescaling is calculated as the difference between the observed condition variable value and the optimal condition reference.
          By normalizing multiple condition variables, different indicators can be more directly compared;
      3. Condition Index Account: combines all indicators together using a weighted mean. Currently, all indicators take the same weight, summing to 1 (e.g., 0.25 when four condition metrics are selected). In future releases of ARIES for SEEA, users will be able to assign custom weights to the indicators to better reflect their local importance when accounting for ecosystem condition.
      
      
      
      

      
      

    • Ecosystem services account (physical terms)

      These accounts measure the biophysical quantities of services provided by ecosystems and used by economic units. Use tables are not explicitly supplied with the model outputs, but use is described in the automatically generated reports for the selected accounts. In the current version, four ecosystem services are available. A fifth one, Nature-based tourism, is in its final stages of development and will be made available in a future ARIES for SEEA release.
      
        
      
      

    • Ecosystem services account (monetary terms)

      These accounts measure the monetary value of the selected ecosystem services, applying SEEA EA-compliant valuation method(s). Use tables are not explicitly supplied with the model outputs, but use is described in the automatically generated reports for the selected accounts. In the current version, three ecosystem services are available. A fifth one, Nature-based tourism, is in its final stages of development and will be made available in a future ARIES for SEEA release.
    • 
      
      
      
      

    • Asset accounts

      Compilation of asset accounts is not supported in the current version, but is planned for inclusion in the future. Asset accounts can be manually compiled by the user by calculating the net present value of monetary supply and use accounts as described in the SEEA EA manual (https://unstats.un.org/unsd/statcom/52nd-session/documents/BG-3f-SEEA-EA_Final_draft-E.pdf).
      
      
      
      

    • Spatial and temporal aggregation

      The user can select how the results of the accounts are aggregated in accounting tables.
      Currently, only the first option is available. In future ARIES for SEEA releases, the user will be able to generate accounting tables as follows:
      
      
      
      1. "Primary only": a single table summarizing results for the entire context identified by the user;
      2. Administrative subregions: multiple tables grouped by subnational jurisdictional entities (i.e., administrative level 1 - State, Province, or District);
      3. Protected areas: multiple tables grouped for protected areas found in the analysis context;
      4. River basins: multiple tables grouped results by watersheds found within the analysis context.
         
         

    • Temporal accounting

      When a multiple-year analysis is selected, the user can decide to output tables:
      
      1. For the first and last years of the time series only, or
      2. For all years within the time period selected.

III. Key outputs and observations

This section shows the most relevant spatial inputs and outputs in the account(s) run by the user
(e.g., in the case of ecosystem extent, the most relevant outputs are land cover and ecosystem type data).

The last section stores all tables and maps produced in that session, so that the user can download them in a zipped file (of Excel spreadsheets or GeoTIFFs) before leaving the application.



In the upper left corner of the application, there is a section dedicated to SEEA-relevant indicators.

This section includes selected indicators from the Sustainable Development Goals (SDGs) and the Convention on Biology Diversity (CBD) Post-2020 Biodiversity Indicators, which have been added to the application.

Once it is opened, a drop-down menu shows the list of available indicators. 

Those selected by the user are added in the ARIES for SEEA panel as if they were an additional standard SEEA EA account

IV. Quick tips

• Set your context (e.g., Cape Town in South Africa). Suggested way:
  
  1. Zoom and pan using the Map boundaries option:
     
  2. Make sure the screen contains your target context:
       
  3. Once the screen contains the area of interest, switch to Administrative region:
      
     
     
  4. If the whole administrative region does not fit the screen, a locator map in the lower-right corner of the interface
     will show the entire context as currently selected:
     
  5. Refine your search further (zooming in/out) to make sure the desired region is selected.
     
     
     
  6. Note: the region automatically selected is not always the most central, but part of the screen occupying the largest area:
     
  7. The name of the geographic entity is displayed on the upper left side of the application.
     
     
     
• When the gear starts spinning, the system is computing the model(s) requested by the user:
  

• To offer the best visualization of the context in analysis, the search bar/results box can be positioned where the user prefers.
  To do so, point the cursor over the search bar, and once this symbol appears, drag it to a preferred location.
  
  
   
   

V. Explore results and methods used to run your accounting models

Aspects of the user interface change once you begin to run an accounting model:

1. Run an accounting model (e.g., ecosystem extent account - simple net balance).
   1. Once the extent account is selected, the search bar will turn yellow.
      A rotating gear will appear next to the selected account(s), which indicates that the model is being compiled.
      When a model input or intermediate output has been computed, it is listed and can be explored individually, by selecting the specific layer (e.g., aridity):
      
   2. When a main output of the model is computed, this is listed (in the darker section) above the other inputs/intermediate output:
      
      
   3. If the model is dynamic (i.e., is calculated for >1 year), the progress bar at the bottom of the search bar/results box progresses as the computation continues:
      
      
      
   4. In the right-upper corner of the app, three tabs are now available:
      
      
      1. 1. Data view (always available)
         2. Documentation view 
         3. Data flow view
            
            
   5. In our example, the application automatically takes the user to the section showing the accounting result(s).
      This happens every time a final outcome is generated:
   6. 
      
      
   7. To review other sections of the documentation:
      
      1. 1. The report
            1. 
               The tree menu on the left facilitates navigation. Each report includes:
               
               1. A general introduction to the model(s);
               2. Information on the SEEA framework or any other more general modeling frameworks (when part of a larger set of models);
               3. The methods applied;
               4. A summary of the main results;
               5. Caveats or other considerations in interpreting model results, as part of the discussion; and
               6. Reference(s) for data and method(s) used.
                  
                  

         2. Tabular results

            
            To ease the observation of the results:
            1. 1. The first column is visible while scrolling to the right,
               2. Results can be sorted by the element in the first column, or by ascending and descending order of the result of any column,
               3. Text size can be adjusted using the buttons in the upper right (with yellow circles).
                  
                  

         3. Image(s) section

            Figure(s) and (maps) generated during the computation are listed in this section and can be explored individually.
            If results come from multi-year models, the maps are interactive and show changes over time.
            
            
            
            

         4. Resources section
            
            

            Lists all the resources used in the computations, to provide full-transparency on the final output.
            Each of the data resource can be explored individually
            
            
         5. Dataflow section: shows a list containing a visual summary of each model component and how these were combined to obtain the final results. In the next version of ARIES for SEEA, this section will also report the decisions taken by the system on:
            1. 1. Which combination of data and models was implemented,
               2. Which data and model alternatives could have been used instead and,
               3. The AI parameters used to support the choices made.
                  
                  
                  

   8. The third tab available once a model has run is the Dataflow view

      This shows a diagram that visually summarize each model component and how those were combined to obtain the final result(s). Clicking on an individual rectangle provides details about the dataset, algorithm, etc. used at each step of the model.                      

VI. Spatialization of statistical aggregates

This subsection explains how ARIES spatializes tabular national statistical data and aggregates it through spatial modelling, using crop provisioning as an example. The problem can be framed as follows:

1. National statistical aggregates provide national-level crop production data in a particular year.
2. For SEEA EA accounting, we also need to map production levels.
3. Spatial data (maps) identifying production of most cultivated crops are available, but often for a limited number of years, not in time series, and at coarse spatial resolution. Such data are typically generated using subnational agricultural production data correlated spatial data (e.g., for land cover or crop locations).

The method implemented follows:

1. 1. 1. Spatially explicit crop production data (3) identify where production for a certain crop occurred in a given reference year (e.g., 2010);
      2.  We overlay spatially explicit crop production data (3) with national boundaries to calculate a crop production reference table for each country (1) and crop. The reference table (1) enables us to: 
         1. Account for changes in political borders; 
         2. Disaggregate production coming from overseas departments/regions included in national statistics (3); and
         3. Reliably produce results at scales aside from national scale (i.e., subnational, supranational)
      3. The spatialization of the production coming from statistical aggregates (4), is thus based on the information on (1), (2) and (3), as shown in the graphic below:

      4. Example of spatial disaggregation: crop production
         
         

(1): Production table (reference)
(2): Production data from statistical aggregates (in reference year)
(3): Crop production map (spatial data in reference year)
(4): Production data from any year
Note that 2 and 4 must come from to the same data source (apply the same methodology), in order to generate spatialized crop production in time series.
The models only estimates the reallocation of such production, the biophysical production comes from statistical aggregate 

This approach assumes crop extent to be constant (based on spatial data for the reference year), with only crop yield (tons produced per hectare) changing over time - an assumption we know to be false, but that a lack of time series spatial data based on subnational crop production forces us to accept, for now. Increases or decreases in production are thus spread across the whole country, rather than for that specific (but unknown) areas. The approach can be improved significantly with:

1. More up-to-date spatially explicit data on crop production; and
2. Statistical aggregates from higher-level administrative units (i.e., at state, region, province, or district level).
3. Such approaches can enable the use of methods that produce more accurate spatial disaggregation of any statistical aggregates. 

VII. Other ARIES features

While most of the functions needed to compile SEEA accounts are available through ARIES for SEEA interface, the normal functionalities of the ARIES Explorer remain available through the search bar.
This sections shows how to query them.

1. Search the knowledge bar (1)

   This box is used to call models, select your context, modify default settings, and show the help page.
   
   
   
   
   

   1. Query an observable from the space bar (A):
      Please note: unlike the search bar in the ARIES for SEEA application, the general ARIES Explorer search bar is case-sensitive.
      
      

      1. Press the space bar on the empty box to receive suggestions.
         
         
         
      2. Start with an upper-case letter to search for a geographical context (invalid if you have already set a context).
         
         
         
      3. Start with a lower-case letter to search for an observation (e.g., elevation),
         
         
         Option (iii) enables users to compute a model available in the ARIES Explorer environment and appropriate for the SEEA framework, but which is not contained as an option in the application.
         This option can be used to explore a wide range of different models (termed observables), which can refer to model inputs or outputs. 
         
         
   2. Other settings/options/help (B)
   3.  
      

      1. Context:

         1. Reset a previously used context
         2. Draw a new context (using your cursor)
      2. Space and time:
         1. Adjust the spatial and/or temporal resolution of your analysis
      3. Options
         1. Activate the interactive mode to add manually add input to the model (adjusting parameter values)
      4. Settings
         1. Remember location (remember the last context selected)
         2. Remember docked status
      5. Help
         1. Show the help tutorial
            
            
   4. New option to show the coordinates of your cursor (they change as you move it around)
      
      
      
      

2. Zoom (2)

   Zoom in (by clicking on the ) and out (). This can be useful when selecting your context using map boundaries (entire rectangle shown in the interface) or administrative regions or river basins, which change as a user zooms and pans the map.

   
   
   

3. Background map (3)

   Select the background map you prefer. By default, ARIES uses OpenStreetMap as it currently offers the best user experience even at higher resolutions.
   
   The user can switch to other available background maps as desired.
   
   

VII. Coming soon

Additional features that will be made available in future releases of ARIES for SEEA include:

1. Various improvements to ecosystem condition accounts:
   1. Option to select multiyear averages rather than single opening and closing years
   2. Custom weighting of metrics for ecosystem condition index accounts
   3. Handling presence/absence and categorical data in ecosystem condition accounts (e.g., burned area)
2. Incorporation of other SEEA-related indicators, such as those from the SDG and CBD frameworks
3. Physical and monetary supply-use accounts for nature-based tourism
4. Reporting by administrative subregions, protected areas, and watersheds as ecosystem accounting areas
5. Various "interactive mode" options for user to supply model inputs and parameters
6. Improved integration of user-supplied spatial data inputs through a drag-and-drop interface
7. Various improvements to functionality and user experience

X. Training Programs and material

XI. About ARIES

• Technical support
• The ARIES team
• ARIES for SEEA funding
  
  

Need to add a final page with an about section where we acknowledge EU funding. 

May be also add a contact (email address) such as support@integratedmodelling.org