RMS Medium Term Perspective on Hurricane Activity

Similar documents
2013 ATLANTIC HURRICANE SEASON OUTLOOK. June RMS Cat Response

3/5/2013 UNCERTAINTY IN HURRICANE FREQUENCY: HOW MODELERS APPROACH THE PROBLEM PROBLEM STATEMENT

Assessment of strategies to project U.S. landfalling hurricanes. Judith Curry

PRMS WHITE PAPER 2014 NORTH ATLANTIC HURRICANE SEASON OUTLOOK. June RMS Event Response

TROPICAL CYCLONES IN A WARMER WORLD

2004 Hurricane Season: Climate Overview and Lessons Learned

A pragmatic view of rates and clustering

April Forecast Update for North Atlantic Hurricane Activity in 2019

Colorado State University (CSU) Atlantic Hurricane Season Forecast

Understanding Climatological Influences on Hurricane Activity: The AIR Near-term Sensitivity Catalog

Tropical Cyclones and Climate Change: Historical Trends and Future Projections

5 Year Prediction of the number of Hurricanes which make U.S. Landfall

SIXTH INTERNATIONAL WORKSHOP on TROPICAL CYCLONES

At the Midpoint of the 2008

Climate Change and Hurricane Loss: Perspectives for Investors. June By Karen Clark and John Lummis

The 2014 Atlantic Hurricane Season. What is New and What to Expect. Mark Chambers President & CEO ImpactWeather, Inc.

Reply to Hurricanes and Global Warming Potential Linkages and Consequences

Colorado State University (CSU) Atlantic Hurricane Season Forecast

Comments by William M. Gray (Colorado State University) on the recently published paper in Science by Webster, et al

FORECAST OF ATLANTIC SEASONAL HURRICANE ACTIVITY AND LANDFALL STRIKE PROBABILITY FOR 2018

July Forecast Update for Atlantic Hurricane Activity in 2016

July Forecast Update for Atlantic Hurricane Activity in 2017

Hurricanes and Climate Change: Expectations versus Observations

Agricultural Outlook Forum Presented: February 17, 2006 THE SCIENCE BEHIND THE ATLANTIC HURRICANES AND SEASONAL PREDICTIONS

April Forecast Update for Atlantic Hurricane Activity in 2018

Assessment of strategies used to project U.S. landfalling hurricanes

July Forecast Update for North Atlantic Hurricane Activity in 2018

Global warming, U.S. hurricanes, and insured losses

April Forecast Update for Atlantic Hurricane Activity in 2016

Low frequency variability in globally integrated tropical cyclone power dissipation

FORECAST OF ATLANTIC SEASONAL HURRICANE ACTIVITY AND LANDFALL STRIKE PROBABILITY FOR 2015

August Forecast Update for Atlantic Hurricane Activity in 2012

The Atlantic Hurricane Database Reanalysis Project

Weather Outlook for Spring and Summer in Central TX. Aaron Treadway Meteorologist National Weather Service Austin/San Antonio

ATMOSPHERIC MODELLING. GEOG/ENST 3331 Lecture 9 Ahrens: Chapter 13; A&B: Chapters 12 and 13

Pre-Season Forecast for North Atlantic Hurricane Activity in 2018

Macroeconomic implications of natural disasters on small economies

2016 Hurricane Season Preview

Tropical cyclones in ERA-40: A detection and tracking method

PERSPECTIVES ON FOCUSED WORKSHOP QUESTIONS

Catastrophe Modeling in an Environment of Climate Change. AIR Worldwide Corporation

August Forecast Update for Atlantic Hurricane Activity in 2015

A Climatology of Landfalling Hurricane Central Pressures Along the Gulf of Mexico Coast

NW Pacific and Japan Landfalling Typhoons in 2000

August Forecast Update for Atlantic Hurricane Activity in 2016

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM SEPTEMBER 13-SEPTEMBER 26, 2018

The 2009 Hurricane Season Overview

Amajor concern about global warming is the

Preliminary Assessment of 2017 Atlantic Hurricane Season. Phil Klotzbach. Department of Atmospheric Science. Phil Klotzbach

Some figures courtesy of: Chris Landsea National Hurricane Center, Miami. Intergovernmental Panel on Climate Change

Hurricane Risk: Importance of Climate Time Scale and Uncertainty

arxiv:physics/ v1 [physics.ao-ph] 7 Nov 2006

The Outlook for the 2016 North Atlantic Hurricane Season Hurricane Genesis & Outlook (HUGO) Project (April 15, 2016)

A Hurricane Outlook for the 21st Century.

Storm Hazard Assessment for San Pedro/Ambergris Caye, Belize. Statistical Estimation of Hazard Frequency and Intensity

The 2005 North Atlantic Hurricane Season A Climate Perspective

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM SEPTEMBER 1 SEPTEMBER 14, 2015

NOTES AND CORRESPONDENCE. What Has Changed the Proportion of Intense Hurricanes in the Last 30 Years?

AnuMS 2018 Atlantic Hurricane Season Forecast

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM AUGUST 16 29, 2018

HURRICANES AND CLIMATE

Climate Outlook through 2100 South Florida Ecological Services Office Vero Beach, FL September 9, 2014

Variations of Typhoon Activity in Asia - Global Warming and/or Natural Cycles?

By: Jill F. Hasling, Certified Consulting. Houston, Texas. November 2016

Relationship between the potential and actual intensities of tropical cyclones on interannual time scales

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM AUGUST 4-17, 2015

Topic 3.2: Tropical Cyclone Variability on Seasonal Time Scales (Observations and Forecasting)

July Forecast Update for Atlantic, USA and Caribbean Landfalling Hurricanes in 2001

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM SEPTEMBER 15 SEPTEMBER 28, 2015

The Outlook for the 2016 North Atlantic Hurricane Season Hurricane Genesis & Outlook (HUGO) Project (June 15, 2016)

AnuMS 2018 Atlantic Hurricane Season Forecast

TSR TROPICAL STORM TRACKER LAUNCH

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM SEPTEMBER 27-OCTOBER 10, 2018

G 3. AN ELECTRONIC JOURNAL OF THE EARTH SCIENCES Published by AGU and the Geochemical Society. Nonlocality of Atlantic tropical cyclone intensities

Hurricane Forecasting Using Regression Modeling

A RETROSPECTIVE ON 10 YEARS OF MODELING HURRICANE RISK IN A WARM OCEAN CLIMATE

WEEKEND WEATHER OUTLOOK BELIZE, CENTRAL AMERICA

Research on Climate of Typhoons Affecting China

SEASONAL CLIMATE FORECASTING TO BENEFIT BUSINESS AND SOCIETY

4. Climatic changes. Past variability Future evolution

11/17/2017. Memorable Hurricanes of 2004 MET 4532

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM OCTOBER 13 OCTOBER 26, 2017

Forecasting U.S. Hurricanes 6 Months in Advance

A Statistical-Dynamical Seasonal Forecast of US Landfalling TC Activity

Observed Climate Variability and Change: Evidence and Issues Related to Uncertainty

Comments on: Increasing destructiveness of tropical cyclones over the past 30 years by Kerry Emanuel, Nature, 31 July 2005, Vol. 436, pp.

Global Warming, the AMO, and North Atlantic Tropical Cyclones

Outlook 2008 Atlantic Hurricane Season. Kevin Lipton, Ingrid Amberger National Weather Service Albany, New York

2015 AIR Hurricane Season Preview

Careful, Cyclones Can Blow You Away!

An Objective Algorithm for the Identification of Convective Tropical Cloud Clusters in Geostationary Infrared Imagery. Why?

SST and North American Tropical Cyclone Landfall: A Statistical Modeling Study

AnuMS 2018 Atlantic Hurricane Season Forecast

Atlantic Hurricanes and Climate Change

Visibility network of United States hurricanes

Variations of Typhoon Activity in Asia - Global Warming and/or Natural Cycles?

and Weather extremes Tropical cyclones Today and in the Future Greg Holland c3we.ucar.edu

QUALITATIVE DISCUSSION OF ATLANTIC BASIN SEASONAL HURRICANE ACTIVITY FOR 2019

AnuMS 2018 Atlantic Hurricane Season Forecast

Weathernews Looks Back at 2015: A Year of Storms. Powerful Typhoons Occurred with Changing Patterns in El Niño Years

Transcription:

RMS Medium Term Perspective on Hurricane Activity Dr. Manuel Lonfat Florida Commission on Hurricane Loss Projection Methodology Workshop Tallahassee, July 27 2006

Agenda Multiyear autocorrelation of Atlantic, US and FL hurricane activity over the last century Appropriate insurance industry time horizon for assessing a forward looking perspective on activity RMS medium term perspective methodology Summary: future plans 2

Issues with Atlantic and US Hurricane Datasets The datasets used in this analysis are HURDAT and the US list of US Hurricanes, both available at NHC The datasets are known to be incomplete ATL basin data likely incomplete prior to 1950 or possibly later regarding the storm intensity The US landfall dataset also has issues, in particular with the max intensity of low Rmax hurricanes before 1900 What is the chance that Charley would have been classified Cat 4 in 1860? Source: http://www.aoml.noaa.gov/hrd 3

Atlantic Basin Hurricane Activity has shown a Marked Increase Since 1995 Since 1995 # CAT 3-5 storms x 2 the 1970-1995 average Principal driver: low latitude sea surface temperatures (SSTs) Competing theories as to the cause of the current high levels of activity: multidecadal oscillation vs. climate trend However statistically and climatologically there is no doubt that we are currently in a persistent period of higher than average activity Annual number of Category 3-53 hurricanes in the Atlantic basin between 1900-2005 and 5 year running mean 4

NOAA s Perspective (May 2006) Atlantic hurricane seasons exhibit prolonged periods lasting decades of generally above-normal or below-normal activity. Hurricane seasons during 1995-2005 have averaged 15 named storms, 8.5 hurricanes, and 4 major hurricanes, with an average ACE index of 179% of the median. NOAA classifies nine of the last eleven hurricane seasons as above normal, and seven as hyperactive. In contrast, during the preceding 1971-1994 period, hurricane seasons averaged 8.5 named storms, 5 hurricanes, and 1.5 major hurricanes, with an average ACE index of only 75% of the median. One-half of these seasons were below normal, only three were above normal (1980, 1988, 1989), and none were hyperactive. 5

Hurricane activity in Florida Cat 1-5 in FL have gone through phases of high and low activity The frequency of Cat 1-5 from 1970 to 1994 was significantly lower than that of the long term. Frequencies of Cat 1-2 and Cat 3-5 have both been above the long term over the last 11 years Since 1995, 7 Cat 1-2 have impacted Florida, giving rise to an annual rate similar to that prior to 1970 Number of FL Cat 1-5 4 3.5 3 2.5 2 1.5 1 0.5 0 1 12 23 34 45 56 67 78 89 100 Years since 1900 Annual number of Category 1-51 hurricanes in Florida between 1900-2005 6

Climate and Hurricanes - Science timeline Pre 2005 The recent increase in Atlantic hurricane activity: Causes and implications - Goldenberg et al. 2001 Penetration of Human-Induced warming into the world s oceans Barnett et al. 2005 Gray 2005 on natural cycles and ENSO Uncertainty in Hurricanes and global warming - Trenberth 2005 Increasing destructiveness of tropical cyclones over the past 30 years Emanuel 2005 2005 Chan 2005 on West Pacific activity trends Landsea 2005 response to Emanuel 2005 Pielke et al. 2005 on trends damage Low frequency variability in globally integrated tropical cyclone power dissipation Sriver and Huber 2006 Trends in global cyclone activity over the past twenty years (1986-2005) - klotzbach 2006 Changes in tropical cyclone number, duration and intensity in a warming Environment Webster et al. 2005 Hurricanes and Global Warming - Pielke 2005 Gray 2005 on data quality prior to satellite era 2006 Atlantic Hurricanes and natural variability in 2005 - Trenberth and Shea 2006 decomposition of contributions to activity Sea Surface temperatures and tropical cyclones in the Atlantic Basin Michaels et al. 2006 Deconvolution of the factors contributing to the increase in global hurricane intensity. Hoyos et al 2006 Hurricanes and Global Warming Potential Linkages and Consequences Anthes et al 2006 Atlantic Hurricane Trends Linked to Climate Change Mann and Emanuel 2006 Blue global warming linkage Agenda Setting Green AMO linkage 7

What is the appropriate time horizon for assessing hurricane risk for the insurance industry? 8

The Appropriate Risk Horizon for CAT Models The long term historical activity rate is inconsistent with the current level of activity For capital management and insurance regulatory purposes, shifting to a short term seasonal risk horizon (i.e. less than 12 months) would prove far too volatile Would imply significant changes in Florida insurance rates year on year (and month on month?) 9

The Appropriate Risk Horizon for CAT Models In September 2005 RMS determined that in future model upgrades the expected activity rates would be for a rolling five year forward looking period More stable than seasonal forecasts Appropriate time interval for scientific forecasting relative to multidecadal variability To be updated annually 10

RMS Medium Term Perspective Methodology 11

Expert Elicitations RMS methodology uses a range of statistical analyses and an elicitation of leading experts in the field Elicitations have been widely used across fields where the length of the record is too short or the return periods too long to use traditional statistical methods: Eg. Assessment of return periods of earthquake fault rupture in California For hurricanes, this approach is relevant because Different theories exist regarding the causes of current high levels of activity, yielding different perspective on the future Issues with the data reduce the confidence in employing long term statistics 12

1 st Expert Elicitation Bermuda, October 15, 2005 The elicitation was organized to obtain a consensus over the activity of the next 5 years (2006-2010) US and Caribbean Formal process (following rules of elicitation) The following questions were asked: Expected annual mean activity rate of Cat 1-5/Cat 3-5 hurricanes in the Atlantic Basin for the period 2006-2010 Expected annual mean activity rate of Cat 1-5/Cat 3-5 hurricanes making one or more landfalls along the US coast in the period 2006-2010 How long will the current period of higher activity last, and how low will hurricane activity then fall? 13

Expert Elicitation of Medium Term Caribbean and U.S. Rates Experts present at the meeting were: Prof. Jim Elsner (FSU), Prof. Kerry Emanuel (MIT), Tom Knutson (NOAA/GFDL) and Prof. Mark Saunders (UCL), reflecting a broad cross section of expertise Experts were provided with detailed data and US (NHC) landfall statistics Range of statistics derived from the time series were also provided, along with description of the way that activity rates are considered in the RMS model A RMS analyst was present at the meeting to perform calculations on the data as requested by the experts 14

Expert Elicitation of US Medium Term Activity Rates Experts activity increment targets were obtained by assessing probabilities of exceedance of key historical activity rates, under the assumption of a Poisson distribution The range of answers was 2.12-2.32 for CAT15 and 0.88 0.94 for CAT35 Two approaches were used for calculating the mean rates across the experts A simple average A weighted average (weights for each expert inversely proportional to the range of their answers) The results from the 2 methods are within 2% or less from each other Experts targets for the next 5-year activity at US landfall are: 21% increase in the activity of CAT12, 36% increase in the activity of major hurricanes (CAT35) compared to the long term mean (1900-2005) Experts all agreed high period of activity likely to last for at least next decade 15

Implementation of activity rate increases Type 5 (Gulf) Type 3 Type 4 Caribbean Type 2 Type I Type I Marginal 20 N 10 N 0 Activity rate increases predominately in Type 1 and Type 2 Atlantic origin hurricanes 16

RMS Climate Hazard Models From May 2006 in all RMS models the default activity is that of the medium term 5 year perspective Recommended to be used for all standard applications of the model as used by insurers, reinsurers, rating agencies, Cat bond issuers & investors, regulators etc 17

Summary: Future Plans on developing Medium Term Hurricane Activity Rates Medium Term 5 year activities will be updated annually Procedures for end 2006 (2007-2011) update well underway to include increased component of multiple statistical and climatological inputs as well as expert elicitation Procedures and outcomes will be submitted for peer-review publication 18