ARIES for SEEA Explorer

The Statistics Division of the UN Department of Economic and Social Affairs (UN DESA) and the UN Environment Program (UNEP), in collaboration with the international research and innovation platform Artificial Intelligence for Environment & Sustainability (ARIES) at the Basque Centre for Climate Change (BC3), have developed a pioneering Artificial Intelligence (AI)-powered application for rapid natural capital accounting: the ARIES for SEEA Explorer!

ARIES is a networked software technology that redefines sustainability assessment and valuation for decision-making. The ARIES approach to mapping natural capital, natural processes, human beneficiaries, and service flows to society is a powerful new way to visualize, value, and manage the ecosystems on which the human economy and well-being depend.

A better way to model ecosystem services

ARIES maps the agents of provision of ecosystem services (sources), their beneficiaries (use), and any biophysical features that can deplete service flows (sinks) automatically choosing the best available models and data. Through artificial intelligence and innovative semantic modeling, ARIES assembles spatial data and expert-contributed model components – deterministic or probabilistic – to quantify and map ecosystem services, at the appropriate spatial scales and specifically for each ecological and socio-economic context.

An ecosystem of models

Models and data used in ARIES are stored on an expanding semantic web. While users can provide their own data and models, an extensible network hosts data, models and model services that are assembled according to context. In this kind of collaborative modelling:

  1. Models and data are developed by the individual, independent experts;
  2. Open source technology allows researchers and institutions to add models and data to the network;
  3. Models and data can be made available to all users or restricted communities;
  4. Artificial Intelligence negotiates the most appropriate models to solve user queries;
  5. Transparent documentation can be generated for all the models used and their integrated results.

Two-step rapid assessment

On the user side, ARIES enables a simple, two-step ecosystem service assessment process. Client software (desktop & soon web-based) allows modeling with minimal configuration and training. After users are logged into the network (through a secure certificate), modeling reduces to

  • Step 1: set the context (for example, searching for a watershed name, or drawing a region on an online map)
  • Step 2: find a concept representing the issue of interest using free-form search, and drop it on the context to observe it.

At that point, ARIES uses its internal knowledge and the data and models on the network to creates agents and processes, connect them into a flow network, and produce a complete assessment of socio-ecological structure and function in the context.

An adaptive modelling technology

Keeping it simple, but realistic

ARIES is an artificial intelligent modeller rather than a single model or collection of models. ARIES chooses ecological process models where appropriate, and turns to simpler models where process models do not exist or are inadequate. Based on a simple user query, ARIES builds all the agents involved in the nature/society interaction, connects them into a flow network, and creates the best possible models for each agent and connection. The result is a detailed, adaptive, and dynamic assessment of  how nature provides benefits to people.

Modelling on the semantic web

ARIES is based on the radically novel k.LAB technology, which allows models and data to be contributed by independent researchers, hosted on a network, and automatically assembled into model workflows following a user’s simple. The technology, which can be applied beyond the field of ecosystem services, is the first operational example of semantically integrated, distributed, collaborative modelling. As an international network of scientific collaborators contribute data and models, the ARIES system grows by itself, and each new assessment automatically adopts the best data and models available. In the next year, users will be able to not only develop, but also run models directly from the world wide web, enabling a simple, two-step modeling workflow suitable for non-technical users, such as decision makers and their staff.