Return to site

Understanding and projecting changes in extreme precipitation events

Author: Thomas R Karl, L.H.D.

· Risk analytics,Science

Why Extreme Precipitation Events Matter

Extreme precipitation is a driving factor in much of the worldwide socio-economic losses including property damage and loss of life. This is supported by data collected both by Munich Re and NOAA. NOAA finds that since 1980 more than 1 trillion dollars in insured and uninsured property losses occurred from U.S. disasters with the most costly accompanied by extreme precipitation (Table 1). Munich Re’s data span the globe and their global assessment indicates that worldwide insured and uninsured property losses from flood-related catastrophes exceed a trillion dollars since 1980 (Munich Re 2017). Munich Re (2018) reported that in 2017 flooding accounted for 49% of all the catastrophic economic losses. It is noteworthy that neither NOAA’s nor Munich Re’s statistics include indirect losses attributed to lost work-days or lost revenue from business closures.

Table 1. Billion-dollar events to affect the U.S. from 1980 through 2016 (CPI-Adjusted). Source NOAA-NCEI with those related to extreme precipitation events highlighted in blue

What is known today

Heavy and extreme precipitation events are increasing across many land areas especially in the mid and high latitudes. Changes in heavy and extreme precipitation were first documented by Iwashima and Yamamoto (1993) who used the data from scores of stations in Japan and the United States. Subsequently, Karl and Knight (1998) were the first to show that heavy precipitation events were increasing both in intensity and frequency across much of the United States. Since then, many analyses of regional and global changes of precipitation extremes have found that extremes of precipitation are increasing and this characteristic has been pervasive as the world has continued to warm into the 21st Century e.g., Groisman et al (2005), IPCC (2013), Alexanderson (2016). Kunkel (Fig. 1) provides a very recent analysis of trends in precipitation extremes across the globe where there are adequate observations to discern trends of 1 in 10-year 5-day extreme precipitation events over the period 1951-2014.

Figure 1. The 1951-2014 trend of number of 5-dy total precipitation events exceeding the station-specific threshold for an average 10-yr recurrence interval. White dots indicate that the trend is significant at the p=0.10 (small dots) or p=0.05 (large dots) level (Interim Progress Report, Kunkel, SERDP project number: RC-2517, U.S. Dept of Defense).

In general, as the global atmospheric temperature increases the amount of water vapor that the atmosphere can sustain increases. This is borne out by observed trends (IPCC, 2013) over the past several decades. With more water vapor in the atmosphere there is the potential for heavier precipitation. For example, Karl and Trenberth (2003) showed that for the same amount of annual average precipitation, those locations with the warmest temperatures (and therefore highest atmospheric water vapor) have more precipitation occurring in the heaviest precipitation events. Assuming atmospheric carbon dioxide concentrations and other greenhouse gases continue to increase at their present rate, global average temperatures will rise by at least 1.5⁰C, and possibly significantly more, e.g., 4⁰C. All climate models predict that this increase will lead to extreme precipitation events changing at a rate much faster than changes in the mean precipitation, especially for North America, many parts of Europe and Asia. In some places, such as parts of the Asian monsoon region, even greater increases are projected, but decreases in parts of the subtropics (Fig. 2 top) are also projected. The rate of extreme precipitation increase is driven in large part by the increase in water vapor and is projected to be larger for the mid-, and especially, high-latitudes of the northern hemisphere as the hottest month of the year (the month that is associated with the greatest amount of atmospheric water vapor) increases more in those regions compared to areas farther south (Fig. 2 bottom). Climatologists know, based on a well-known equation (the Clausius–Clapeyron equation) that expresses the relationship between water vapor and temperature, that the average rate of precipitation increase should be constrained to be about a 6 percent increase per one-degree Celsius increase of global temperature. This excludes however, changes in atmospheric circulation, including the thermodynamics of the atmosphere, (e.g., those big summer thunderstorms), which can substantially alter this relationship.

Figure 2. (top) Fractional changes (%) of precipitation and the maximum amount of water in the atmosphere as projected by seven state-of-the-science climate models (CMIP5). These are multi-model mean differences (future minus present) in the 30-year period maximum daily precipitation for 2071–2100 under a business as usual emissions scenario (RCP8.5) scenario, relative to the 1971–2000 reference value. (bottom) A scatterplot of grid point differences (future minus present) of the 30-year maximum amount of water in the atmosphere versus 30-year average temperature of the climatologically warmest month in the lower atmosphere (850hPa) for 2071–2100 with respect to 1971–2000 for the business as usual emissions scenario (RCP8.5). The straight line represents a slope of 6.3%/⁰C (from Kunkel et al., 2013).

Although the greatest quantity of atmospheric water vapor is in the tropics (Fig. 3 top), because greater warming is expected in the high latitudes, the fractional increase in water vapor is greater in the high latitudes (Fig. 3 middle). An important modifier of the effect of more atmospheric water vapor is atmospheric vertical velocity (Fig. 3 bottom). Ascending air enhances extreme precipitation events, while descending air curtails these events. Understanding the causes of the changes in atmospheric motion is critical to understanding regional variability of the expected changes of extreme precipitation. This is a key focus of new research being undertaken by a few members of The Climate Service’s Advisory Board in a separate research study. It is briefly described in the next section.

Figure 3. (top) The 30-year maximum water in the atmosphere for 1971–2000 (mm), averaged over seven climate models (CMIP5). (middle) Fractional changes (%) of the maximum water in the atmosphere (PWmax) and (bottom) vertical atmospheric motion (ωmin) projected by the same seven climate models. These are multi-model mean differences (future minus present) of the 30-year maximum values under the business as usual scenario (RCP8.5), for 2071–2100 relative to the 1971–2000 (middle and bottom) (from Kunkel et al., 2013).

Researching new knowledge

As shown by Mall et al. (2007) in regions where the atmospheric circulation changes little, climate models project strong agreement between increases in extreme precipitation events and temperature with the sole consideration being the Clausius–Clapeyron equation, but there are many areas where this is not strictly followed. This is because of the impact of changes in atmospheric circulation on extreme precipitation events. A key research issue is to understand the causes of the changes in atmospheric circulation. This understanding will enable improved confidence in both regional and seasonal variability of extreme precipitation events (see Pfahl et al., 2017 for some seasonality considerations).

In new research being carried out by members of the Advisory Board and their colleagues there are several areas of atmospheric circulation changes being analyzed in context with changes in water vapor and extreme precipitation events. This includes: (1) changes in frontal frequency e.g., warm, cold, stationary, (2) changes in extra-tropical cyclones, (3) changes in tropical cyclones, and (4) changes in monsoonal circulation. Each of these factors is being examined with respect to both the observed climate record and the projected change in climate models. The former is to gain a better understanding of the causes of recent regional variability of changes in extreme precipitation. The latter is to see which climate models are consistent with known causes of observed changes in precipitation extremes, and to better understand the changes in the regional variability of extreme precipitation events within the models.

Kunkel et al., (2013) have already laid the groundwork for analyzing atmospheric circulation in context with extreme events. They found that extreme precipitation events are associated with various atmospheric circulation systems. For the U.S. they found that the percent of daily extreme precipitation events were primarily associated with well-known circulation features including: fronts, extra-tropical cyclones, tropical cyclones, and monsoonal circulations.

Changes in frontal frequency

A major research challenge for the team working on understanding and projecting changes in extreme precipitation events was developing an automated procedure to identify fronts. Today, fronts are still ‘hand-analyzed’, but this is impossible to do when examining scores of climate model simulations with a hundred or more years of data. Fortunately, using tools of machine learning a satisfactory machine-based method has been developed (in peer-review), and is now being applied to both historical data and climate model output. This will enable a better understanding of the causes of climate model changes of extreme precipitation in context with known relationships with frontal frequency and extreme precipitation.

Changes in Extratropical Cyclones

Another key factor associated with precipitation extremes, extratropical cyclones, is being examined with respect to both changes in storm tracks and storm intensity. Climate models project changes in both, and understanding how these changes affect changes in the intensity and frequency of local and regional extreme precipitation events will be useful for assessing confidence in those projections.

Changes in Tropical Cyclones

Tropical Cyclones account for over ten percent of the extreme precipitation events in the U.S. and they are expected to change in both frequency and intensity across the globe. However, a major impediment to using this information to understand and project changes in regional extreme precipitation events has been inadequate spatial resolution of climate models. Fortunately, newer models being used in the analysis have spatial resolutions better than 25 km, and some as low as 5 km.

Changes Monsoonal Circulation

As mentioned in the previous section some of the largest changes in extreme precipitation are projected in monsoonal circulation systems. Analyses of the changes in strength of the monsoon are being explored to better understand how these evolve in climate models in relation to the observed climate. Those models that are able to demonstrate changes in extreme precipitation commensurate with changes in monsoonal strength will be identified. Ultimately, this will lead more confidence in some models.

Putting it all together

An important goal will be for the first time to weight those climate model projections more heavily which are most consistent with our understanding of why precipitation extremes change. Today, we are just beginning to understand which climate models work best for particular problems, in contrast to considering them all equally likely representations of the climate system. This ‘coming-of-age’ of climate model projections is not dissimilar to what has already occurred with weather models. The meteorological community generally understands which weather models perform best during certain weather types and for specific regions. So, our ability to discriminate among the many climate model projections for changes in regional and local extremes of precipitation will ultimately be commensurate with our ability to discriminate among weather model predictions for specific areas and weather.


Groisman P. Y., R.W. Knight, D.R. Easterling, T. R. Karl, G.C. Hegerl, V.N. Razuvaev, 2005: Trends in Intense Precipitation in the Climate Record. Journal of Climate, 18, 2326-1350.

Alexander, L.V., 2016: Global observed long-term changes in temperature and precipitation extrmes: A review of progress and limitations in IPCC assessments and beyond. Weather and Climate Extremes , 11, pp. 4-16

IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G –K Plattner, M. Tignor, S.K Allen, J. Boschung, A Nauels, Y. Xia, V. Bex, and P. M Midgley (eds.)] Cambridge University Press, Cambridge, United Kingdom and New Your, NY, USA, 1535 pp, doi:10.1017/CB09781107415234

Iwashima, T., and R. Yamamoto, 1993: A statistical analysis of the extreme events: Long-term trend of heavy daily precipitation. J. Meteor. Soc. Japan, 71, 637–640.

Karl, T.R. and R.W. Knight, 1998: Secular Trends of Precipitation Amount, Frequency, and Intensity in the United States.  Bulletin of the American Meteorological Society, 79, No. 2, pp 231-241.

Karl, T.R. and K.E. Trenberth, 2003: Modern Global Climate Change. Science, 5, pp. 1719-1723. DOI 10.1126/science/1090228

Kunkel, K.E., D.R. Easterling, D.A.R. Kristovich, B. Gleason, L. Stoecker, and R. Smith, 2012: Meteorological causes of the secular variations in observed extreme precipitation events for the conterminous United States. Journal of Hydrometeorology, 13 (3), 1131-1141, doi:10.1175JHM-D-11-1-8.1,

Kunkel, K.E., T.R. Karl, R. Easterling, K. Redmond, J. Young, X. Yin, and P. Hennon, 2013: Probable maximum precipitation and climate change Geophysical Research Letters, 40, 1402–1408, doi:10.1002/grl.50334, 2013

Mall, P., M.R. Allen, D.A. Stone, 2007: Testing the Clausius–Clapeyron constraint on changes in extreme precipitation under CO2 warming Climate Dynamics March 2007, 28, pp 351–363

Munich Re, 2017 Geo Risks Research, NatCatSERVICE


Munich Re, 2018


Pfahl, S., P. A. O’Gorman, E. M. Fischer, 2017: Understanding the regional pattern of projected future changes in extreme precipitation. Nature Climate Change, DOI: 10.1038/nclimate3287

All Posts

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!

OKSubscriptions powered by Strikingly