Top Banner
Towards Automation of Model Execution from a Decision Support Environment MODSIM 2011, Perth, Australia Ralf Denzer 1
22

Towards Automation of Model Execution from a Decision Support

Mar 14, 2022

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Towards Automation of Model Execution from a Decision Support

Towards Automation of Model Execution from a Decision Support Environment

MODSIM 2011, Perth, Australia

Ralf Denzer

1

Page 2: Towards Automation of Model Execution from a Decision Support

Outline

� Project objectives� Users� Approach� Common Services� Scenario Management System� Conclusions

Towards Automation of Model Execution from a Decision Support Environment

2

Page 3: Towards Automation of Model Execution from a Decision Support

Objectives

� Climate change

� What are the impacts on cities?

� How can urban planners consider the effects ofclimate change in their planning processes?

SUDPLAN is developing a 3D/4D web-basedplanning, prediction and training tool to supportlong term urban planning including the effects ofclimate change

3

Page 4: Towards Automation of Model Execution from a Decision Support

Climate Change Infusion

4

Regional ClimateModel

Urban DownscalingServices• Rainfall• Hydrology• Air Quality• Climate Scenarios

Local Model

Page 5: Towards Automation of Model Execution from a Decision Support

Objectives of the EDSS

5

Model Control

Data Management

Visualisation

Integration

Reporting

Regional ClimateModel

Urban DownscalingServices• Rainfall• Hydrology• Air Quality• Climate Scenarios

Local Model

Page 6: Towards Automation of Model Execution from a Decision Support

User Categories

Analyst

System

ManagerModeller

Reporting &

Decision Making

Service Provision

& Customizing

Modell

Development

Scientific

Workflows

Scenario

Management

Page 7: Towards Automation of Model Execution from a Decision Support

User Categories

Analyst

System

ManagerModeller

Reporting &

Decision Making

Service Provision

& Customizing

Modell

Development

Scientific

Workflows

Scenario

Management

Support decision makers, expert planners, city

planners

Page 8: Towards Automation of Model Execution from a Decision Support

Users Categories

Analyst

System

ManagerModeller

Reporting &

Decision Making

Service Provision

& Customizing

Modell

Development

Scientific

Workflows

Scenario

Management

Vision and Requirements of Scenario-driven Environmental DSS Supporting Automation for End Users

develop, integrate, configure mathematical

models

Page 9: Towards Automation of Model Execution from a Decision Support

Users Categories

Analyst

System

ManagerModeller

Reporting &

Decision Making

Service Provision

& Customizing

Modell

Development

Scientific

Workflows

Scenario

Management

Vision and Requirements of Scenario-driven Environmental DSS Supporting Automation for End Users

Host, provide, install & maintain, administer

EDSS instances

Page 10: Towards Automation of Model Execution from a Decision Support

Approach

Provide 2 Building Blocks� Services providing climate

change information on an urban scale

�Common Services

� Scenario Management Platform to support tasksinvolved in urban planning

�Scenario Management System

10

Page 11: Towards Automation of Model Execution from a Decision Support

Common Services

� Provide climate scenario data on a European scale, available for every city. Provide models capable of downscaling data to city level.

� All models and results are accessible through OGC service interfaces (SWE )

� City specific input improves the accuracy of the results

11

Functionality

� Rainfall time series and IDF downscaling� Storm water generator� Hydrological conditions and runoff downscaling� Air quality downscaling

Page 12: Towards Automation of Model Execution from a Decision Support

Scenario Management System

� Highly interactive,3D/4D graphics-based, DSS-environment

� Users define, manage, execute and explore different decisions and simulate decision scenarios

� Visualisation, comparison and documentation of different decisions

Page 13: Towards Automation of Model Execution from a Decision Support

Requirements for the SMS

� Usability, Interactivity� Automation� Profiling� Information Management� Scenario Support (what If)� Model Management, Calibration & Validation� Advanced Visualisation� Documentation, Sharing, Publishing� Platform Management & Integration Support

13

See: Schlobinski et al. in ISESS-2011,SUDPLAN Deliverable D3.1.1 Requirements Specification

Page 14: Towards Automation of Model Execution from a Decision Support

Common Data

ApplicationDomain

System Feature

Local Model Integration

Common Service Integration

AdvancedVisualisation

Comparison

Climate ∆ Precipitation

Air Pollution Trends

Climate ∆ Temperature

Climate ∆ Air Quality

Urban Drainage

Urban Wastewater

Migration Trends

Climate ∆ Hydrology

Climate ∆ EffectsTraining

Scenario Management

SUDPLAN Dimensions

ApplicationLogic

ModelLogic

GenericLogic

SpecificLogic

Page 15: Towards Automation of Model Execution from a Decision Support

Extension and Customisation Mechanisms

15

generic functionality

provide customization

Page 16: Towards Automation of Model Execution from a Decision Support

Realisation Status

� Model management (for Common Services) including management and basic visualisation of model input and output data (2D)

16

Page 17: Towards Automation of Model Execution from a Decision Support

Realisation Status

� Interactive visualisation of spatio-temporal data

17

Page 18: Towards Automation of Model Execution from a Decision Support

Realisation Status

� 3D visualisation of the terrain model and gridded results from local air pollution model

18

Page 19: Towards Automation of Model Execution from a Decision Support

Benefits (Common Services)

� Urban planners can easily include climate aspects in their planning processes

� Using the models they are using anyway� Without the need to develop a specific model

cost(integration) << cost(development of specific model)

19

Page 20: Towards Automation of Model Execution from a Decision Support

Benefits (SMS)

� Platform for EDSS that supports a large number of tasks

cost(customization) << cost(development)

� Can be transferred to other applications� Comes with integrated Common Services

20

Page 21: Towards Automation of Model Execution from a Decision Support

Summary

21

We aim to ...

Improve the integration of climate changeinto urban planning decisions

Improve the link between climate change research, local modelling and urban planning

Increase the efficiency of systems development

Page 22: Towards Automation of Model Execution from a Decision Support

22

SUDPLAN is an EU FP7 funded projectunder the Information CommunicationTechnology program (ICT-2009-6.4),running 2010-2012.

Contact

Coordinator: Lars Gidhagen, [email protected]