STANDARDS FOR MODELLING OF FLOODING IN OPEN COASTS AND LARGE ESTUARIES

Transcription

1 STANDARDS FOR MODELLING OF FLOODING IN OPEN COASTS AND LARGE ESTUARIES HAKEEM JOHNSON CH2M, UK ASGHAR AKHONDI-ASL, IAIN ANDREWS Environment Agency, UK Flood & Coast 2017 Conference

2 Outline of presentation Motivation Objectives Model Standards Assessment of model quality National baseline of coastal/estuary flood model standards Treating models as assets Conclusions 2

3 Motivation 150 billion assets & 2.5 million people at risk from coastal flooding in the UK. Coastal flood models are important because they are used to: Underpin key investment decisions to manage coastal flood risk. provide flood information for emergency response. provide the evidence to prevent inappropriate development in the flood plain. Existing models are not designed to nationally consistent standards. Consequences are: Lack of clarity of what is good enough Model quality can vary widely No systematic approach to model maintenance a challenge with limited resources. 3

4 Personal Story Client : I want a coastal flood model at minimum cost. Cumbria Me (modeler) : How much effort should we be put into this? Effort = Cost. What is good enough for this project? Source Standards clarifies what is required for both parties. 4 Dawlish 2004

5 Objectives Develop nationally consistent standards for coastal flood modelling. Good Enough Target standard Efficiencies Develop & maintain to target standards Future Investment Confidence in evidence used in decisions National baseline of model standards for existing coastal flood models. Understand standards of existing models ID Areas of Improvement Plan of model Improvements Understand existing evidence ID improvements to target standard Plan for the future 5

6 Model standards 6

7 Model standards Target quality standard based on intended use/s. It defines what is good enough. Model quality standard measures how well key flooding processes (Sources, Pathways and flood spreading model to Receptors) are represented in the model. Model condition standard measures how up-to-date a model is, in terms of key data, technology, flood events and changes to capital schemes. Overall model score based on model quality and condition. 7

8 Target quality standard Model quality scores Intended use/s Detailed Scheme design (eg crest levels, flows, volume) Individual Property level protection Real-time inundation & real-time threshold crossing (could include preprepared maps), and Flood Warning areas for Higher Risk communities Appraisal of flood risk management solutions Damage estimation based on depth (AAD/EAD) Real-time threshold crossing (could include pre-prepared maps), and Flood Warning areas for Lower Risk communities National Flood Risk Assessment (NaFRA) [input to quality levels A or B] Coastal cell - Broad scale investment information [Strategy Report] Broad scale information for Flood Guidance Statement Weighted annual average damage (WAAD) estimation Flood Alert area & Forecast scenarios in the day T day period (exclude real-time inundation maps). Unsuitable for flood modelling at the current state of knowledge and technology. 8

9 Model standards Model technical quality Is a measure of how well key flooding processes are represented in the model. Model condition Is a measure of how up-to-date a model is. Conceptual understanding how current? Conceptual understanding how current? Source data quality tides, surges, waves, river discharge etc. Changes to Data: Topo surveys, hydrographic surveys, Capital schemes Pathway data quality - topo, bathy, defences, breaching etc. Changes to Methods hydrology / hydraulic methods 9 Model build quality - type, resolution, validation etc. Scoring is based on evidence obtained from literature and current practice. Occurrence of major flood event. Scoring is based on evidence from publically available GIS databases.

10 Model technical quality Coastal/Estuary Flood Model Quality Standard Quality control and modelling processes Conceptual Understanding using S-P-R framework Data on SOURCES of flooding Data on Flow PATHWAYS Model build properties & reliability for predicting flooding at RECEPTORS Modeller skill and experience 10

11 Assessment of model quality 11

12 Assessment of model technical quality Technical Quality Score SPR Source Pathway Model Build Elements Elements Elements Element score Component score Technical Quality score 12 Element score is based on evidence from literature or practice. Component score = weighted scores for all elements. If any critical element score is U, score = U. Quality score = average of component scores. If any component score is U, score = U.

13 Assessment of model technical quality Component 0 Component 1 Component 2 Component 3 Conceptual understanding Source DATA Flow Pathway DATA Model Build Event Combinations Bathymetry Inundation model - type Elements within each component Climate Change Coastal Water Level Beach Morphology Flood Defences IM mesh resolution IM validation Wave Overtopping Volumes Topography IM - sensitivity OS maps 13

14 Assessment of model technical quality Element scoring example Flood inundation model - type A: 2D depth averaged models. Examples: TUFLOW, MIKE 21 Evidence: Defra/EA, Benchmarking the latest generation of 2D hydraulic modelling packages. FLOODsite Consortium, Evaluation of Inundation Models, Report T , Revision Number 1_7_P15, April 2007 B: As in (A) C: Simplified 2D model (2D-) OR GIS level or volume contour or 1D bucket model subject to caveats. U: None of the above OR caveats for "C" not met. 14 Caveats: Flood plain < 350m, <100 receptors & no essential infrastructure. GIS level - no defences.

15 Assessment of model technical quality Element scoring examples Surge Profile A: Average profile from 15 storms Data from Class A gauge OR CFB method B: Average profile from 10 or more storm surge events Evidence : Environment Agency (2011) Coastal Flood Boundary Conditions for UK mainland and islands - Project: SC060064/TR2: Design sea levels, Flood and Coastal Erosion Risk Management Research and Development Programme, February Experience from practice - Project Team C: Average profile from 5 or more storm surge events. U: Effect of storm surge not considered or considered incorrectly 15 The effect of the storm surge profile may be quite critical in some cases, but this is difficult to quantify without site specific sensitivity testing.

16 Assessment of model technical quality Element score Component score Model Quality score Element score is based on evidence from literature or practice. Component score = weighted scores for all elements. If any critical element score is U, score = U. Quality score = average of component scores. If any component score is U, score = U. 16

17 National Baseline Study Baseline study on: Existing Environment Agency coastal/estuary models 17

18 Model Score Code RA Technical Quality (RATQ) Model Condition (MC) Score Description Score Description 0 A/B Model score meets or exceeds target quality score of B 1 A- / B- Meets or exceeds target quality score (breach analysis not considered). 2 C / C- Model score below target quality score 3 U Below standard. Model needs major improvements to bring it up to date. Good / Very Good Fair Poor Very Poor Model meets or exceeds target condition of Good 1 step below target condition 2 steps below target condition 3 steps below target condition. Brighton beach Model Overall Score = Max (MC Score code, RATQ score code) 18

19 NATIONAL BASELINE STUDY 53 45% 0 0% 3 3% 25 21% 10 8% 6 5% 0 0% 24 20% 26 22% 26 22% Model quality A A- B B- C C- U Model condition Very Good Good Key numbers 16 to 17 No of elements checked in quality scoring 119 No of existing coastal and estuary flood models audited with new standards 42% Percentage of models that meets target quality (B/B-) 47% Percentage of models that meets target condition (Good) % 30 25% Fair Poor Top 4 areas for model improvements are: 1) Breach analysis; 2) LiDAR data; 3) Bathymetry data and 4) Tide data. Baseline presented at 1-day workshops to 14 Environment Agency area teams.

20 Model quality (RATQ) score: National 20

21 Model condition score: National 21

22 A new way of managing models Treating models as assets 22

23 A new way of managing models Develop model to target quality Efficiency, Value for money & Consistency Good enough! Agree target model quality Check model quality and condition Health & Safety Asset management cycle Clarify Objective Specify model use/s Carry out model maintenance Sustainability 23 Retire asset & develop new model

24 Summary & Conclusions 24 Model standards have been developed to guide management of coastal flood models. Target standard, Model quality & model condition Benefits of managing models with new standards Clarity on what is good enough Efficiency and national consistency Greater confidence in evidence Model maintenance culture National baseline of coastal/estuary flood model standards has been carried out. 42% of the 119 models met the target quality of B/B- 47% met the target condition of Good Framework includes work on real-time models

25 Way forward At the Environment Agency: Coastal flood modelling standards is being incorporated in ITT for projects The standards are being applied in ongoing projects: Eg. Outer Thames estuary wave modelling project Baselining study is being used to guide future model improvements It will be good to extend use of the standards to the wider industry. 25

26 Cumbrian coast Source Dawlish 2004 Acknowledgements: We would like to acknowledge the contributions of : From the Environment Agency: Project Team: Neil Ryan, Niall Hall, Tim Hunt, Helen Colyer, Jonathan Boyes and Vicki Dutchburn; Various Area Teams that provided data and attended workshops for the National baseline study; From CH2M: Project Team: Caroline Stuiver, Luke Lovell, Andy Parsons, Karim Rakha, John Scrase and Laura Baird. 26

27 Thank You Crosby Beach.