Summary Of Modelling PDF

Summary of Modelling **Introduction lecture** A model is a simplification of some other structure or a system; of a "real" phenomenon Wicked systemic problems? - Problems lacking clarity regarding scope, aims, solutions and potential feedbacks and problem-shifting - There is no number o...

Summary of Modelling **Introduction lecture** A model is a simplification of some other structure or a system; of a "real" phenomenon Wicked systemic problems? - Problems lacking clarity regarding scope, aims, solutions and potential feedbacks and problem-shifting - There is no number of solutions or approaches to a wicked problem; all of them are unique - Models help simplify this complexity, exploration of scenarios Mathematical modelling and systems thinking: - Explicit models are needed: assumptions are laid out, others can replicate the results - Clearly define and conceptualize your model aims and scope - Clearly think through model limitations and excluded aspects Possible roles and functions of quantitative models: - Testing hypotheses, scenario analysis and prediction, Participative Modelling etc. Cohort-competent Models: - Demographic models used to project population changes over time - Specific cohorts are age, sex or other demographic factors - Allow for a more nuanced understanding of population dynamics and are used for public planning purposes Key components include: 1. Fertility: number of births in each cohort over time 2. Mortality: number of deaths in each cohort 3. Migration: movement of individuals in and out of the population - They help create projections about future population sizes and structures Exponential growth models: assume a population grows at a constant rate without considering age structure or limiting factors Logistic function models: introduce carrying capacity suggesting population growth will slow once it approaches its limit but does not account for age structure Time-delayed logistic function: similar to logistic functions but incorporates delays in the response to changes in population density (still lacks detailed cohort analysis) Lotka Volterra Equations: used for modelling predator-prey interactions and do not focus on age structure or demographic components MISO Model Multiple inputs, single output - Analyze systems where various factors contribute to a single result - How different inputs interact and influence the outcome Exogenous variables: determined outside the model; not influenced by other variables within the model but can affect model outcomes. Example: government policy, weather conditions, international market trends Endogenous variables: determined within the model; values influenced by other variables within the system; main focus of analysis they are the outcomes or results of a model. Example: consumer demand, price levels, production output Socio-Ecological Metabolic Modelling Analyzes material and energy flows through the socio-economic systems, focusing on metabolism-like processes. Analyzes how resources are converted within the system. Example: Input-Output flows: resources entering and leaving a system; Strengths: - Broad coverage of flows - Holistic view of socio-ecological transitions Weaknesses: - Requires extensive data (quantitative data on material and energy flows across systems) - Assumes system linearity Philosophy: - Rooted in material balance principles (total input equals total output, accounted for storage, transformation and losses). These align with the law of conservation of mass (matter cannot be created nor destroyed) - Focus on sustainability and transitions for efficient resources use Application: - Resource efficiency studies - Environmental impact studies Insights and Limits: Provides insights of resource use dependencies but lacks dynamic feedback loops or agent behaviours. Scenario: A city´s government wants to evaluate its resource usage and waste generation to transition toward a circular economy. By analysing the flow of materials, energy and emissions within the city´s boundaries this model identifies hotspots of inefficiency, such as high energy consumption in transportation and excessive waste from construction. **Session 2:** A system is a group of interacting or interrelated elements that act according to a set of rules to form a unified whole. - It is described by its boundaries, structure, and purpose and it expressed in its functioning Systems can be: - Isolated: no exchange of energy nor matter - Closed: exchange of energy but not matter - Open: exchange of both energy and matter Complicated Systems Complex Systems --------------------------------------------------------------------------- -------------------------------------------------------------- Elements have a minimal degree of independence from each other Strong dependencies between elements Removal of an elements does not fundamentally change the systems behavior Removal of an element leads to profound change (destruction) Predictable and reducible Non-reducible Systems thinking consists of elements, interconnections and function / purpose. System Dynamics Modelling Explores temporal dynamics of systems using stocks, flows and feedback loops to capture interdependencies over time. Strengths: - Captures feedback loops and delays - Effective for long-term policy analysis Weaknesses: - Requires simplifying assumptions (to make them more computationally manageable) - Example: total population rather than individual agents (focus on large-scale behavior) - May simplify complex systems to achieve tractability Data: - Estimates from historical data - Historical data - Qualitative interviews - Workshops on data collection - High level aggregated data on key system components, interdependencies and temporal patterns Philosophy: - Systems thinking (looking at the whole, the relationships rather than just the parts) - Focus on capturing dynamic interdependencies Application: - Policy design and analysis - Scenario analysis in economic, environmental and energy systems Insights and limitations: - Good for analysing temporal changes but struggles with granular detail or agent-specific dynamics Scenario: A policymaker wants to understand how introducing a carbon tax will affect emissions and economic growth over 50 years. Feedback loops show how taxation affects industrial behaviour, public demand and long-term emissions reductions. Applies to dynamic problems -- problems that change over time -- any dynamic system characterized by interdependence, mutual interaction, information feedback and circular causality - Focuses on the dynamic interrelations between stocks and flows It is a methodology for studying complex dynamic systems that include [nonlinearities, delays and feedback loops. ] Central concepts: - Stocks (accumulation): they have a certain value at a time and can change over time; ex: bank account, population etc. - Flows (change over time): changes of a stock per time step; ex: births, deaths, payments to and from an account - Feedback loops (circular causality): a system´s output can also be that system´s input Feedback determines the [dynamic process of a system] and [how things evolve over time. ] 1. Reinforcing (positive): self-reinforcing loops. Example: More population leads to more births, more births leads to a bigger population which leads to greater births and so on the direction of change is the same. Component A increases, so does component B. 2. Balancing (negative): self-correcting. Example: As populations grow, death rates will be higher than what it would normally be, more deaths leads to a smaller population than what it would normally be direction of change is opposite. Component A increases, therefore component B decreases. Steps of SDM: 1. Causal loop diagram: a visual presentation of key variables and how they are interconnected. It focuses on the variables and the causal relationships between them (which variables, are they connected and how -- positive or negative?) 2. Stocks and Flow diagram: define units, stocks and flows Validation: - Structural tests -- testing the plausibility of the model - Behavioural tests -- can the model reproduce the dynamics of the system - Sensitivity tests -- tests of variables that introduce uncertainties into the modelling Quality of the model depends on [system understanding, available data and validation procedures. ] **Session \#3** Basic climate processes: - Solar radiation - Absorption (gas, aerosols) - Reflexion (clouds, aerosols) - Absorption and reflexion from the surface - Geometry: orbit, angle of axes, rotation of earth Fundamentals of numerical weather prediction (NWP) - Forecasting weather conditions using mathematical models that simulate the behaviour of the atmosphere Key concepts: 1. Governing equations: NWP is based on fundamental equations of fluid dynamics and thermodynamics, which describe atmospheric process and motion 2. Initial conditions: accurate forecasts depend heavily on high-quality initial conditions derived from data sources and data assimilation Climate Modelling Simulates Earth´s climate systems to study changes over time, including energy balances, temperature changes and emission pathways. Strengths: - Accurate for physical and chemical processes - Critical for climate scenario projections Weaknesses: - Uncertainty in predictions - Challenges in modelling socio-economic drivers Data: - Meteorological, hydrological and geophysical data - Emissions inventories - Energy balance data Philosophy: - Physically-based modelling - Focus on integrating atmospheric, oceanic and land processes - Explicit integration step by step from now into the future "prognostic modelling" - Need a "closed system" base model has to be global Application: - Climate change impact assessments - Emissions scenario testing Insights and limitations: - High resolution climate insights but limited in integrating socio-economic dynamics Scenario: A regional government needs to predict how increased GHGs will influence precipitation patterns and temperature. The model predicts severe droughts and storms, prompting adaptation strategies for agriculture. - Provides critical data for climate adaptation and mitigation strategies To use climate models for future climate projections, we need models that realistically represent all relevant processes and forcings. - Climate models have to be able to reproduce the observed development (example: temperature anomalies) and assumptions about what is changing in the future. - Example: scenarios about different paths to reach certain atmospheric conditions. **Session \#4** Characteristics of complex social systems: 1. Networks of heterogenous social actors 2. Nonlinearity 3. Different time scales 4. Delays and accumulation of stocks 5. Resilience 6. Balance of power and narratives The standing ovation problem - To sit or to stand equilibrium will be reached quickly Why SOP? - Social dynamics are complex - Macro-behaviour emerges from micro-motivations - Spatiotemporal dynamics Social systems cannot be adequately simulated with linear mathematical models, while ABM offer a way to examine their dynamical properties. Agent-based modelling Simulates behaviours of individual agents and their interactions to understand emergent phenomena in complex systems. Strengths: - Captures heterogeneity of agents - Flexible and dynamic for multiple contexts Weaknesses: - Computationally intensive - Difficult to validate or calibrate for large systems Data: - Data on agent attributes, rules on behaviour and interaction dynamics - Requires conceptual understanding of agent heterogeneity Philosophy: - Bottom-up approach - Microlevel interactions leading to macrolevel outcomes Application: - Urban development - Market simulations - Behavioural dynamics modelling Insights and limitations: Provides granular insight into agent behaviour but can be computationally demanding and challenging to validate. Scenario: Retail company that wants to simulate consumer purchasing behaviour when introducing a new eco-friendly product. Agents with diverse preferences and incomes simulate market dynamics, showing how pricing and marketing affect adoption rates. Identify objects (agents) and their decision-making Agents: persons, households, vehicles, products and companies Identify interactions between objects and between objects and their environment - The results are the aggregated behaviour of the whole systems - Allows spatial localisation Simply put: ABMs model the behaviour of agents Dynamics of the system emerge from the interaction processes (bottom-up) Agents can be based on qualitative (preferences, values, or strategies) or quantitative (indicators weighted by value judgements) outputs. How valid is it to use models based on simplification to understand real world social dynamics? - Most valuable aspects of social sciences cannot be captured by a formal tool - However, emergence can occur even in binary systems, in which complexity arises from simplicity - Most important variables are those that qualitatively change the dynamics of a system, and often those are quite few Typologies for Social Ecological Systems are based on: 1. Functional roles: e.g. household, farmer, politician 2. Preferences: the decision structure is the same in each role but subtypes are formed by preferences 3. Behavioural mechanisms: e.g. risk-averse-risk taking, imitation, deliberation, repetition. Scaling -- 3 possibilities 1. Scaling out: same model, extended data set, larger geographical unit. - Advantages: no new model necessary - Disadvantages: data availability, increasing processing time, adaptability to larger geographical units 2. Scaling up: aggregate agents into institutions. - Advantages: little data required (compared to scaling out) - Disadvantages: representability of interactions (do institutions behave like individuals?); need to remodel process 3. Nesting (multi-model approach): nesting of individual agents at institutional level; visualising interactions and reactions: governance regimes to individual behaviour, support for policy makers, typologisation as an important component Cellular automata (CA) - Spatially discreet systems - CA compute interactions of cell states that can, with the right rules, create dynamic emergent phenomena - Can be used to model complex social behaviour - Easily translates into geographical maps (GIS) - Easily allows spatial representation of social entities (agents) **Session \#5** **Environmentally-Extended Input-Output Model** Analyses inter-industry flows of resources and emissions within and across economies, extended to include environmental impacts. Strengths: - Strong data consistency due to balancing principles - Effective for assessing global supply chain impacts Weaknesses: - Static framework - Strong assumptions on product/price homogeneity - High demand for sectoral disaggregation Data: - Inter-industry transaction tables - Data on material and energy flows, emissions and resource use - National and global economic statistics Philosophy: - Duality of physical and monetary flows - Focus on balancing input-output relations across sectors Application: - Carbon and ecological footprint analysis - Environmental accounting - Supply chain assessment Insights and limitations: Robust insights on interdependencies but lacks temporal dynamics and flexibility for non-linear systems. Scenario: A Carbon Footprint analysis looks at: a. Territorial emissions: from production and consumption b. Carbon footprint: emission anywhere on the planet, embodied in products and services Static models: no temporal dynamics, no feedback loops The economy is dual and circular: - Duality: goods and services flow between two agents - Products and monetary flow Circularity results from specialization / division of labour, every seller of one product buys many other products systems of interdependencies with many direct and indirect effects Balancing principles: foundational accounting rules used to ensure consistency in modelling systems 2 compatible reports on 1 event or symmetric matrix (avoids double-counting by recording inter-entity exchanges only once, maintaining consistency across the system). - Ensure inputs match outputs for every trade flow between 2 sectors Double book-keeping: One activity´s outputs are another´s inputs. Entries are simultaneously inputs and outputs. Balanced economic system: Total output equals total demand. Total expenditures equal total income. The Leontief Inverse - Calculating all inputs required to deliver one unit of a specific type of final demand, ex: a train - Accounts for all interactions between manufacturing sectors Methodological assumptions: - Leontief Production Function: a linear production function - Homogeneity of product / sector outputs and homogenous prices for all deliveries - Price-Quantity equivalence: homogeneity of monetary and physical inter-industry interactions Product mix and price homogeneity: - Unrealistic assumptions that each sector delivers one given mix of products and services at a uniform price - Results in aggregation bias if assumption is validated but can be reduced through finer sectoral disaggregation Domestic Tech Assumption: assume that imports were produced based on the same economic structure as comparable domestically produced goods. Conclusion: - Balancing inputs and outputs of sectors using matrix algebra - Basis of GDP calculations - Describes structure of the economy and allows assessment of direct and indirect effects - Can allocate resource inputs and environmental effects to final demand, disaggregated by sectors / products - Methodological and data problems **Sessions \#6** **Participative Modelling** Involves stakeholders in model development to ensure practicality and inclusivity in representing problems. Strengths: - Builds consensus among stakeholders - Improves understanding and buy-in for complex problems Weaknesses: - Subject to biases from participants - Results may vary depending on group dynamics Data: - Quantitative and qualitative data from stakeholders - Feedback on iterative model structures - Consensus-oriented model adjustments Philosophy: - Co-production of knowledge - Integrating diverse perspectives Application: - Policy development - Collaborative problem solving - Environmental conflict resolution Insights and limitations: Fosters inclusivity and practical insights but may lack technical precision in highly complex scenarios. Objectives: - Gaining a common understanding of a problem - Assisting collective decision-making processes - Explaining implicit knowledge, preferences and values - Improving the legitimacy of a model - Enhance individual and social learning - Informing and enhancing collective action PM process should consider: 1. Reasons and intensions of stakeholders 2. Reasons and intensions of modellers Components of the PM process: 1. Scoping and abstraction: concepts, models, stakeholders, participants 2. Envisioning and goal setting 3. Model formulation 4. Data, facts, logic, cross-checking 5. Model application to decision making 6. Evaluation of outputs and outcomes 7. Facilitation of transparency Actual sequence is adaptable: 1. Issue(s) 2. Needs 3. Methods How are stakeholders selected? - Can never be all-inclusive - Self-selected or invited - Recognized civil society or anyone - Balance between breadth and depth of engagement Stakeholders are mainly involved in: - Providing data for model and model calibration - Evaluation of modelling results (outputs and outcomes) Why want stakeholder participation? - Important local knowledge - Fill in data and information gaps - A participatory process may help mobilize and justify funding - Trust and confidence, transparency for acceptance - PM is a learning exercise between multiple actors Treatment of Uncertainties in PM 1. Evaluation of input uncertainties 2. Uncertainty propagation by models 3. Model outputs and outcomes -- Output uncertainty analysis; adaptive strategies for predicted outcomes and scenarios Participating judgements, decisions and informed actions: a. Human biases: well managed groups and processes needed b. Behaviours: from individuals to groups = human motivation c. Ownership starts with the local, the present and the demonstrable =\> participating approaches viewed as disillusioning, lead to better decision-making and can help take greater ownership of their resources and environments