Trophodynamic (Fach) / Modeling (Lektion)
In dieser Lektion befinden sich 15 Karteikarten
Modelin
Diese Lektion wurde von saha_rass erstellt.
- How to develop a model 1. Model formulation: -make assumptions, choose variables, formulate equations 2.Mathematical solution: solve equations 3. Interprete: interprete the solution but be critical and compare with reality 4.Problem: specifiy problems 1. remodel if necessary
- double doughts when using model as a hypothese evaluation double doughts when used to evaluate hypothese = model correct, hypothesis correct = model correct, hypothesis incorrect =model incorrect, hypothesis incorrect =model incorrect, hypothesis correct
- Modeling 3 Ms 3 M’s : Monitoring, Modeling, Management
- Example halifax aiirport nud daraus gezogene schlüsse = not enough to only monitor eco systems, as unexpected results can occure after a change (eg microclimate change at halifax airport after removal of the trees – always foggy) -Scnearios like this can be prevented when developing a model -Not enough to only extrapolate the past
- application ecopath model 1)adress ecological questions 2)evaluate effects of fisheries on ecosystem 3)Explore management policiy options 4)Evaluate biomass and flow 5)Evaluate environmental changes 6)Analyze placement and impact of marine protected areas
- Ecopath Masterequation Mass balance model consumption = production >>linear model
- Data input ecopath 1) Biomass 2) biomass accumulation 3) Die composition 4)Food connsumption 5) assimilation rate Economical: 1)Landings 2) Fix costs and variable osts 3) discards 4)market price by fleet and group
- ow to develop a model (schematic) MModel formulation make assumptions Set variables and parameters Set mathematical equations Mathematical solution Solve equations InterpretationInterpret solutions!! Be critical of solutions, possibly reformulate the model Compare with reality ProblemSpecify the problem
- WHat are the elements of a model Forcing functions: external variables eg nutrient input, contaminant flow..climate.. State variables: gives information about the ecosystems state usually very obvious eg nutrient concentration, contaminant concentration Mathematical equations: depicts physical, biological and chemical processes in the model ...one process doesnt al2ways have the same equation but different processes can have the same Parameters: Koefficient in the mathematical equations (converters) which are usually given in a range . high potential for mistakes as often lack of knowledge/hard to predict eg growth rate, accumulation rate universal variables: gravity, atomic weight...
- 5 steps after completing the model 1) verification: is the solution realistic? Did we expect this solution? 2) Sensitive analysis: change variables on min 2 levels to see sensiticy in the model (afterwards broad understanding of sensitive nd less sensitive processes in the model) Calibratino: compare model solution with real data eg from observstion or experiments 5) validation: use independent et of data and run the model (rarely done as lack of data)
- Dynamic steady integrated model Dynamic model: ecosystem always changing thus diynamic variables (hard to understand sometime whhats going on) > develop time series >trace evolutionary systems Steady state: variables are constant (not equilibrium) - steady state of system - eg climate forrest birth rate and death rate are constant or energy budget (longterm) integrated model: averages variables over certain time interval. usually very large scale and used globally. eg co2flux, annual production, energy or nutrient budgets
- continous model discrete model continous model: continous time flow eg lotka volterra predetor/prey model: never stops - infinite discrete model: ecological changes occur at fixed time intervalls - finite model ! uses only integer numbers (ganzzahlig)
- Special models PDE IBM spatial modeling Partial differential model: very complex model, rarely used..usually for oceanography Individual based model: describes complex behavior b/w individual s model based on observation of individuals which each posseses individual set of state variables ( physiological traits, spatial distribution, behavior) model acts upon set of rules which are prioritizedmodel acts on a grid layout >cellular automata rarely used as needs huge computers Spatial modeling: difficult to describe organisms movements - last years more development eg. biodiversity model
- Recommendation find balance b/w data, problem,ecosystem,knowledge used wide range of senstivity analysis for best find prameters to find parameters use literature, experiments, observations ,calibration of sub model and entire model models based on databank should not be used for prognotstic models models based on knowledge powerful to set research priorities
- example for ecological model Nutrient cycle eutrophication model global model ecotoxilogy model