Impact Forecasting’s New U.S. Riverine Flood Model: Fresh …events.aon.com/SiteCollectionDocuments/201407_analytics... · 2017. 9. 15. · Aon Benfield | Analytics Proprietary

Prepared by Aon Benfield Analytics

Impact Forecasting’s New U.S. Riverine Flood Model: Fresh Perspective on US Flood Risk Siamak Daneshvaran, Impact Forecasting 2014 Analytics Insights Conference July 22 - 24

Aon Benfield | Analytics

Proprietary & Confidential 2

Agenda

Section 1 Challenges

Section 2 Solutions

Section 3 Benefits

Section 1: Challenges



Challenges

Existing flood tools offer limited underwriting and aggregation capabilities

– Rely on static flood maps which focus on hazard rather than loss

– Do not account for loss correlations across multi-location policies and portfolios

– Deterministic in nature and do not quantify uncertainties

Recent feedback from property underwriters

– “Historical loss ratios show flood risk is not adequately priced”

– “Flood risk is regularly subsidized by other perils”



Issues with FEMA’s incomplete floodplain

FEMA flood extents are limited to 100 and 500 year return periods

FEMA flooding maps do not cover the entire U.S.

– Only main rivers are covered in many areas



Inundation issues

Hydrologic Modeling:

– Transferring precipitation to run-off

– Estimation of inundation and flood footprint using run-off model



Quantifying levee uncertainty

Modeling federally engineered structures



Exposure uncertainty

Building floor level, elevation and basement are key components to flood risk assessment



Agenda


Section 2 Solutions

Section 3 Benefits



Section 2: Solutions



Solutions

Advances in hydraulic data modeling and computational power are supporting new developments in probabilistic flood risk modeling

– Event simulation-based models account for complex spatial and temporal loss correlations

– Efficient modeling of commercial and E&S policy conditions and flood limits

– Reliable flood modeling applications for both underwriting and portfolio aggregation



Impact Forecasting U.S. flood modeling timeline

*Hurricane Sandy insured flood loss estimate includes NFIP loss and Aon Benfield private industry loss estimate

Economic losses in 2014 values (Source: Aon Benfield Impact Forecasting)

2007 2008 2011 2013 2014

Probabilistic U.S. Storm Surge

model introduced

2005 2012

2008 Ohio Flood

Economic loss $16B

2011 Mississippi Valley Flood

Economic loss $4.2B

2013 Colorado Flood

Economic loss $3B

Hurricane Ike

NFIP loss $2.9B Hurricane Katrina

NFIP loss $19.5B

U.S. Inland Flood model

introduced for portfolio

PML modeling

U.S. Storm Surge model

enhancement due Jan 2015

2015

U.S. Inland River Flood model

enhancement for underwriting and

portfolio modeling due Aug 2014

Hurricane Sandy *

Insured flood loss $22.7B



2014 Riverine flood model enhancements

Robust network of stream vectors – 2.3 million km river length

Quality controlled gauge data – focus on river discharge (18,000 gauges)

Cross sectional analysis with hydraulic-based method

Updated DEM data from recent LiDAR collections - Higher resolution elevation data

Greatly expanded event set size with new simulation approach – 75,000 simulated events

Hydrological

Model

Hydraulic

Model

Event

Generation

Intensity

Calculation

Risk Characterization

Damage

Calculation

HAZARD VULNERABILITY LOSS

Policy Conditions

Insured

Loss

Event-based

Flood Extent

EXPOSURE



Two Probabilistic Modeling Approach

Rain Precipitation and Run-off Stream Gauge Observation



Global Circulation Model (GCM)

Advantages:

GCMs can be used for longer term climate change modeling

GCMs can reflect global interactions

Disadvantages:

GCM is very complex

Down-scaling adds uncertainty

Do not account for snowpack accumulation

Uncertainty

Precipitation Model Run-Off Model



Impact Forecasting Uses a Direct Engineering Approach

Stream gauges:

– Provide direct measurement of flow and river height

– Capture all types of flood events

– Errors related to instrumentation and methods are within 5-10%

– Directly estimate correlations

– Generally cover areas with exposure

Less Complex

No Run-off model necessary

Computationally efficient

Error = 5 –

10%



Data sources and model inputs

Important model parameter selections

– Manning’s n values from land-use land, land-cover database

River Network System

Levee System Information

National Elevation Dataset (DEM)

Stream gauges Database

– 1940 - 2011 Records



Hydraulic flood routing

Time Q0 Q1 Q2 Q3 Q4

0.00 0

0.25 8

0.50 28

0.75 54

1.00 88

1.25 133

1.50 187

1.75 232

2.00 263

2.25 280

2.50 283

2.75 280

3.00 263

3.25 244

3.50 221

3.75 193

4.00 159

4.25 130

4.50 110

4.75 93

5.00 79

5.25 59

5.50 42

5.75 30

6.00 22

6.25 16

6.50 11

6.75 8

7.00 6

7.25 4

7.50 3

7.75 1

8.00 0

Time Q0 Q1 Q2 Q3 Q4

0.00 0 0

0.25 8 0

0.50 28 0

0.75 54 1

1.00 88 4

1.25 133 8

1.50 187 15

1.75 232 26

2.00 263 42

2.25 280 61

2.50 283 84

2.75 280 110

3.00 263 136

3.25 244 160

3.50 221 182

3.75 193 200

4.00 159 213

4.25 130 220

4.50 110 222

4.75 93 218

5.00 79 209

5.25 59 197

5.50 42 182

5.75 30 165

6.00 22 148

6.25 16 130

6.50 11 112

6.75 8 96

7.00 6 80

7.25 4 66

7.50 3 54

7.75 1 44

8.00 0 35

Time Q0 Q1 Q2 Q3 Q4

0.00 0 0 0

0.25 8 0 0

0.50 28 0 0

0.75 54 1 0

1.00 88 4 0

1.25 133 8 0

1.50 187 15 1

1.75 232 26 1

2.00 263 42 3

2.25 280 61 5

2.50 283 84 9

2.75 280 110 15

3.00 263 136 22

3.25 244 160 33

3.50 221 182 45

3.75 193 200 61

4.00 159 213 78

4.25 130 220 96

4.50 110 222 115

4.75 93 218 134

5.00 79 209 151

5.25 59 197 167

5.50 42 182 179

5.75 30 165 188

6.00 22 148 193

6.25 16 130 194

6.50 11 112 192

6.75 8 96 187

7.00 6 80 178

7.25 4 66 168

7.50 3 54 155

7.75 1 44 142

8.00 0 35 128

Time Q0 Q1 Q2 Q3 Q4

0.00 0 0 0 0

0.25 8 0 0 0

0.50 28 0 0 0

0.75 54 1 0 0

1.00 88 4 0 0

1.25 133 8 0 0

1.50 187 15 1 0

1.75 232 26 1 0

2.00 263 42 3 1

2.25 280 61 5 2

2.50 283 84 9 4

2.75 280 110 15 6

3.00 263 136 22 10

3.25 244 160 33 16

3.50 221 182 45 23

3.75 193 200 61 33

4.00 159 213 78 45

4.25 130 220 96 59

4.50 110 222 115 75

4.75 93 218 134 92

5.00 79 209 151 110

5.25 59 197 167 127

5.50 42 182 179 143

5.75 30 165 188 158

6.00 22 148 193 170

6.25 16 130 194 179

6.50 11 112 192 185

6.75 8 96 187 188

7.00 6 80 178 187

7.25 4 66 168 183

7.50 3 54 155 176

7.75 1 44 142 167

8.00 0 35 128 156

Time Q0 Q1 Q2 Q3 Q4

0.00 0 0 0 0 0

0.25 8 0 0 0 0

0.50 28 0 0 0 0

0.75 54 1 0 0 0

1.00 88 4 0 0 0

1.25 133 8 0 0 0

1.50 187 15 1 0 0

1.75 232 26 1 0 0

2.00 263 42 3 1 0

2.25 280 61 5 2 1

2.50 283 84 9 4 1

2.75 280 110 15 6 2

3.00 263 136 22 10 4

3.25 244 160 33 16 7

3.50 221 182 45 23 11

3.75 193 200 61 33 17

4.00 159 213 78 45 24

4.25 130 220 96 59 33

4.50 110 222 115 75 45

4.75 93 218 134 92 58

5.00 79 209 151 110 73

5.25 59 197 167 127 88

5.50 42 182 179 143 105

5.75 30 165 188 158 121

6.00 22 148 193 170 136

6.25 16 130 194 179 150

6.50 11 112 192 185 162

6.75 8 96 187 188 172

7.00 6 80 178 187 178

7.25 4 66 168 183 181

7.50 3 54 155 176 182

7.75 1 44 142 167 179

8.00 0 35 128 156 174

Maximum Q0 Create Input

Hydrograph

Muskingum-Cunge

Routing Method

Manning Equation

Next Point (100-meter) Downstream

Wetted Perimeter,

Surface width,

Velocity, Slope and

Roughness

Maximum Qi and Hi

At each Gauge Station



Flood extent methodology



Hazard validation – Colorado flood of 2013

NASA Earth Observatory Impact Forecasting



Event-set generation

Stochastic event set is generated based on smoothing techniques and smearing of historical data (from 1940 to present)

This method preserves the natural correlation existing between the gages in a given event and propagates the historical pattern through the simulated event set considering inter- and intra-event uncertainty

IF simulated event set contains about 75,000 events

Events are generated based on, and are fully compatible with, USGS hydraulic unit codes (HUCs)



Flood of 2008: 41,000

cfs

Flood of 1993: 28,200

cfs

Hazard validation – Upper Mississippi

Station Number Station Name Drainage Area (mi2) Historical Records

05454500 Iowa River at Iowa City, IA 3271 72

Typical gauge discharge return period analysis



Simulated scenarios



Vulnerability functions’ scheme

Basement / foundation are taken into consideration for relevant building classes

Construction Type Stories Coverage Occupancy

Wood 1 Structural Residential

Light Metal 2 Content Commercial

Steel 3 Time Elements Industry

Concrete 4 – 6 (mid-rise) Agriculture

Masonry 7 – 10 (high-rise) Utility

Mobile Home > 11 Education

Vehicle Unknown Transportation

Unknown Government

Unknown



Damage functions validation

1 2 3Claims IF Model 1 2 3

Claims IF Model

1 2 3Claims IF Model

Katrina Ivan

Ike

one story two stories three stories

New damage functions are calibrated based on loss data



Sample AAL and PML Analysis

$-

$2

$4

$6

$8

$10

$12

$14

$16

10 100 1000 10000

Loss

(U

SD)

Bill

ion

s

Return Period (Log Scale)

IF Current Model IF New Model

Primary Zone Return Period IF Current Model IF New Model

California 10000 $ 12.53 $ 14.31

California 5000 $ 12.53 $ 14.02

California 2000 $ 12.48 $ 13.29

California 1000 $ 11.94 $ 12.50

California 500 $ 10.83 $ 11.43

California 250 $ 9.48 $ 10.33

California 200 $ 9.09 $ 9.87

California 100 $ 8.02 $ 8.67

California 50 $ 6.39 $ 7.39

California 20 $ 2.67 $ 5.46

California 10 $ 1.15 $ 3.35

California AAL $ 0.46 $ 0.72

California SD $ 1.50 $ 2.18

$ In billions of dollars



Agenda


Section 2 Solutions

Section 3 Benefits



Section 3: Benefits



Improves underwriting decisions by better differentiating between individual policies

More accurately accounts for loss correlations across multi-location policies and portfolios

Expanded geographic hazard coverage to enable a more complete view of portfolio risk

Benefits



Summary

The 2014 U.S. Riverine flood model:

– Offers significant advancement in assessing flood risk

– Represents a culmination of six years of flood research conducted since the first model version

– Expands the stochastic event-set that improves spatial and temporal coverage

– Improves high resolution LiDAR data for elevation modeling

– Extensively validates vulnerability functions using NFIP and client portfolios



Questions?



Contacts

Siamak Daneshvaran, PhD PE ARe ARM

Global Head of R&D

Impact Forecasting LLC

+1.312.381.5886

[email protected]

mailto:[email protected]





Legal disclaimer

Legal Disclaimer

©Aon Limited trading as Aon Benfield (for itself and on behalf of each subsidiary company of Aon Corporation) (“Aon Benfield”) reserves all rights to the content of this report (“Report”). This Report is for distribution to Aon Benfield and the organisation to which it was originally delivered only. Copies may be made by that organisation for its own internal purposes but this Report may not be distributed in whole or in part to any third party without both (i) the prior written consent of Aon Benfield. and (ii) the third party having first signed a “recipient of report” letter in a form acceptable to Aon Benfield. Aon Benfield cannot accept any liability to any third party to whom this Report is disclosed, whether disclosed in compliance with the preceding sentence of otherwise.

To the extent this Report expresses any recommendation or assessment on any aspect of risk, the recipient acknowledges that any such recommendation or assessment is an expression of Aon Benfield’s opinion only, and is not a statement of fact. Any decision to rely on any such recommendation or assessment of risk is entirely the responsibility of the recipient. Aon Benfield will not in any event be responsible for any losses that may be incurred by any party as a result of any reliance placed on any such opinion. The recipient acknowledges that this Report does not replace the need for the recipient to undertake its own assessment.

The recipient acknowledges that in preparing this Report Aon Benfield may have based analysis on data provided by the recipient and/or from third party sources. This data may have been subjected to mathematical and/or empirical analysis and modelling. Aon Benfield has not verified, and accepts no responsibility for, the accuracy or completeness of any such data. In addition, the recipient acknowledges that any form of mathematical and/or empirical analysis and modelling (including that used in the preparation of this Report) may produce results which differ from actual events or losses.

The Aon Benfield analysis has been undertaken from the perspective of a reinsurance broker. Consequently this Report does not constitute an opinion of reserving levels or accounting treatment. This Report does not constitute any form of legal, accounting, taxation, regulatory or actuarial advice.

Limitations of Catastrophe Models

This report includes information that is output from catastrophe models of Impact Forecasting, LLC (IF). The information from the models is provided by Aon Benfield Services, Inc. (Aon Benfield) under the terms of its license agreements with IF. The results in this report from IF are the products of the exposures modelled, the financial assumptions made concerning deductibles and limits, and the risk models that project the pounds of damage that may be caused by defined catastrophe perils. Aon Benfield recommends that the results from these models in this report not be relied upon in isolation when making decisions that may affect the underwriting appetite, rate adequacy or solvency of the company. The IF models are based on scientific data, mathematical and empirical models, and the experience of engineering, geological and meteorological experts. Calibration of the models using actual loss experience is based on very sparse data, and material inaccuracies in these models are possible. The loss probabilities generated by the models are not predictive of future hurricanes, other windstorms, or earthquakes or other natural catastrophes, but provide estimates of the magnitude of losses that may occur in the event of such natural catastrophes. Aon Benfield makes no warranty about the accuracy of the IF models and has made no attempt to independently verify them. Aon Benfield will not be liable for any special, indirect or consequential damages, including, without limitation, losses or damages arising from or related to any use of or decisions based upon data developed using the models of IF.

Additional Limitations of Impact Forecasting, LLC

The results listed in this report are based on engineering / scientific analysis and data, information provided by the client, and mathematical and empirical models. The accuracy of the results depends on the uncertainty associated with each of these areas. In particular, as with any model, actual losses may differ from the results of simulations. It is only possible to provide plausible results based on complete and accurate information provided by the client and other reputable data sources. Furthermore, this information may only be used for the business application specified by Impact Forecasting, LLC and for no other purpose. It may not be used to support development of or calibration of a product or service offering that competes with Impact Forecasting, LLC. The information in this report may not be used as a part of or as a source for any insurance rate filing documentation.

THIS INFORMATION IS PROVIDED “AS IS” AND IMPACT FORECASTING, LLC HAS NOT MADE AND DOES NOT MAKE ANY WARRANTY OF ANY KIND WHATSOEVER, EXPRESS OR IMPLIED, WITH RESPECT TO THIS REPORT; AND ALL WARRANTIES INCLUDING WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE HEREBY DISCLAIMED BY IMPACT FORECASTING, LLC. IMPACT FORECASTING, LLC WILL NOT BE LIABLE TO ANYONE WITH RESPECT TO ANY DAMAGES, LOSS OR CLAIM WHATSOEVER, NO MATTER HOW OCCASIONED, IN CONNECTION WITH THE PREPARATION OR USE OF THIS REPORT.