E - Changing Return Periods

E - Climate Change Fingerprint

E - Communicating shifting extremes

E - Exceedance Curves

E - Forecasted Probabilities

Small average changes in temperature can hide dramatic changes at extremes

1oC - Socioeconomic impacts of climate change already manifesting

1900-2010 (2011) Anomalies are clearly trending towards hot and very hot temperatures

1900-2010 (2011) U.S. temperature anomalies not nearly as clear as other parts of the world

1950-2010 (2011) Fraction of surface area that's cold or hot shifting significantly

1950-2010 (2011) One can see a systematic warming signal in a number of regions

1950-2010 (2011) The extreme heat tail of anomalies has shifted by 1 SD over 3 decades

1950-2010 (2011) We can track how annual anomalies diverge in terms of SDs from the baseline

1950-2011 (2011) The fraction of land area with very high extremes is increasing

1955-2007 (2011) We can track how surface temperatures are changing over time

1955-2011 (2011) The proportion of surface area displaying >3 SDs is growing significantly

2009 Choosing the right strategy in probability space

2009 Risk of temperature directly related to concentrations, and varies widely over time

2010 Extreme event "tail risks" are substantial

2010 In probability distributions, modes and means and uncertainty are important

2010 rainfall anomaly against future probability distributions

2011 Probability distributions of climate sensitivity

2012 Extreme events are growing in probability faster than projected by the models

2012 Ratio of high to low records in the U.S. 1950's to 2000's

2013 High risks increase faster than mean risks

2013 Probability of catastrophe by treatment

2013 temperatures against future probability distribution

2015 The shock scenario

2016 Annual mean surface temperature anomaly

2016 Changing likelihood of extreme events

2016 Comparison of El Nino strength and temperature trend

2016 Decadal mean surface temperature anomaly

2016 Global surface temperature, annual and running mean

2016 ocean temperature anomaly against future probability distribution

2017 It's all about probabilities

2019 A cumulative distribution function (CDF) transforms a probability distribution into an exceedance curve from 0 to 100%. A complementary CDF (CCDF) does the reverse, from 100% to 0%

2019 Visualizing the long tail of the probability distribution

2030-2050 (2020) Risk of >15% global yield failure increases 2x and 5x

2040 (2013) Median Projected Change in 1% Annual Flood Discharge

2040 (2013) Median Projected Change in SFHA

2050 (2020) Economic impact of extreme flood could be 5-10 times greater than today

2060 (2013) Median Projected Change in 1% Annual Flood Discharge

2060 (2013) Median Projected Change in SFHA

2080 (2013) Median Projected Change in 1% Annual Flood Discharge

2080 (2013) Median Projected Change in SFHA

2100 (2013) Median Projected Change in 1% Annual Flood Discharge

2100 (2013) Median Projected Change in Flood Hazard Parameter

2100 (2013) Median Projected Change in SFHA for 2100

2100 (2013) Monte Carlo Distribution for 1% Annual Flood Discharge change

Abnormal markets - stock exchange not normally distributed

An example of tail risk in the maritime sector

Attribution of changing water scarcity to climate and population

b. By increasing variability with the existing mean

Beta distribution

Brownian motion and stock prices - normal distributions

CCDF complementary cumulative probability function

Change in exceedance probability for normally distributed data

Change in risk of occurrence due to a trend in the mean

Changing global water scarcity risk levels

Changing likelihood of Australian extreme events

Changing likelihood of flooding

Changing temperature probabilities in Colorado River Basin

Conceptual diagram of shifting probabilities

Does climate fit into normal distribution? Evidence that it does not

Effect of mitigation on global emergence in drivers of ecosystem stress

End-century project flow reductions Colorado River Basin

Expected extremes based on linear change in SD

Future warming projections constrasting model vs. paleo estimates

Gamma distribution

Geographic variation in attribution variables

Increase in probability of extremes

Interpreting the IPCC's Results

It's a Matter of Risk and Probabiliy

Lognormal distribution

Mapped Amplification Factors

Mapping flood levels

Mid-century projected flow reductions Colorado River Basin

Moscow July Anomalies

Normal/Gaussian distribution

Northwest Atlantic cod harvest

Power law = better fit for stock extremes

Probability distribution with varying uncertainty around the same mean

Probability distribution with varying uncertainty around the same mode

Reconstructed global mean temperatures

Redrawing this figure using a ratio scale

Regional variation in water scarcity

Relatively modest mitigation takes tail risk off the table

Responding to fire risks: unlikely but we buy fire insurance

SAT Surface Air Temperature

Sensitivity of global SAT to radiative forcing anomalies

Shifting distribution of temperature anomalies in the Northern hemisphere, winter and summer

SST Sea Surface Temperature

Stock market performance on a ratio scale

Storylines used for RCPs and SSPs

Tail risk for maritime accidents

The "tails" of normal and fat-tailed distribution

The distribution of outcomes of rolling 10 dice at once

The distribution of outcomes of rolling two dice

The more dice, the closer you get to a "normal" distribution

The normal probability curve

Three possible outcomes in terms of probability shifts

understanding shifts in extremes due to climate change

Variations in severity of water scarcity

Weibull distribution

Why uncertainty increases risk

2020-2050 (2020) Increase in long tail risk of hurricane damages in Florida

And climate tipping points may be much more likely than we've thought

And there's a clear analogy to abrupt climate change

Beware of long tails that lead to underestimaing risks

Precautionary principle: making decisions without probabilities sensible, but don't take to extremes

The fact is that we spend a lot of money to avoid long-tail risk

What about the venus effect?

What causes extreme events in nature and in finance