Winter weather statistics span several different measures: snowfall, snow cover, temperature, storm damage, health outcomes, and forecast horizons. The figures below focus mainly on the United States and identify the period, population, geography, and measurement definition for each result.
Key Winter Weather Statistics
The most broadly useful figures include:
- In 2023, 1,024 U.S. deaths were attributed to excessive cold or hypothermia using specified death-certificate codes.
- On January 26, 2023, snow covered nearly 60% of the contiguous United States.
- By the end of November 2023, Anchorage had received more than 39 inches of snow.
- By the end of 2023, Anchorage had accumulated more than 78 inches of snow.
- From 1948 to 2021, central-U.S. seasonal snowfall trends ranged from a 110 mm increase to a 180 mm reduction across analyzed weather types and subregions.
- From 1948 to 2021, five of 10 central-U.S. snowfall-producing weather types showed changes partly attributable to event frequency.
- Central-U.S. seasonal snowfall had an average coefficient of variation of about 43% in the long-term climatology summarized by the study.
- The United States recorded two billion-dollar winter storm or cold-wave events in 2024.
- In 2024, U.S. winter-weather disasters caused an estimated $3.6 billion in damage.
- From 1980 through 2024, 24 U.S. winter storm events exceeded $1 billion in losses each in the inflation-adjusted disaster series.
- A January 2024 storm across the central, southern, and northeastern United States caused an estimated $1.9 billion in damage and 41 deaths.
- A January 2024 Pacific Northwest storm caused an estimated $1.7 billion in damage and 20 deaths.
- In 2023, January represented 19.9% of U.S. excessive-cold or hypothermia deaths.
- In 2023, December represented 16.7% of U.S. excessive-cold or hypothermia deaths.
- During 2003–2013, 13,419 U.S. deaths were hypothermia-related under the study’s broader contributing-cause definition.
- During 2003–2013, males accounted for 9,050 hypothermia-related deaths, or 67% of the U.S. total.
- During 2003–2013, mean hypothermia-related death rates were 1.8 per 100,000 for U.S. males aged 65 and older.
- During January–April 2014, 27 hypothermia-related deaths were recorded in Wisconsin surveillance during a record-cold period.
- NOAA’s deterministic Winter Storm Severity Index provides a three-day outlook of expected winter-storm impacts in the United States.
- NOAA’s probabilistic WSSI-P provides a seven-day ensemble-based outlook and was operational from December 2023.
Contents
- Snowfall, snow cover, and seasonal extremes
- Temperature and snowfall trends
- Winter storm damage and losses
- Cold-weather mortality
- Vulnerable populations and exposure
- Forecasting and measurement context
Snowfall, Snow Cover, and Seasonal Extreme Statistics
Snowfall is the amount of snow that falls, while snow cover is the area that has snow on the ground at a particular time. They are not interchangeable: the contiguous United States can have broad snow cover on a snapshot date without that percentage representing a seasonal snowfall total.
On January 26, 2023, snow covered nearly 60% of the contiguous United States, according to NOAA’s 2023 climate assessment. This is an observed area snapshot, not a measurement of total winter snowfall. NOAA, “Assessing the U.S. Climate in 2023”
Anchorage illustrates how a local seasonal accumulation can differ from a national snow-cover snapshot:
| Observation | Period | Result |
|---|---|---|
| Anchorage snowfall | October–November 2023 | More than 39 inches |
| Prior November record in Anchorage | November 1994 | 38.8 inches |
| Anchorage accumulated snowfall | October–December 2023 | More than 78 inches |
The more-than-39-inch November 2023 total exceeded Anchorage’s prior November record of 38.8 inches, set in 1994. The year-end figure exceeded 78 inches by the end of 2023; it is a cumulative total, not a winter-season climatological normal. NOAA, “Assessing the U.S. Climate in 2023”
U.S. Temperature and Snowfall Trend Statistics
Long-term snowfall trends depend on location, storm type, and the period selected. A peer-reviewed central-U.S. study classified 10 distinct snowfall-producing synoptic weather types over 1948–2021, so its results should not be presented as one national snowfall trend.
Across the analyzed central-U.S. subregions, the majority of weather types showed seasonal snowfall trends ranging from a 110 mm increase to a 180 mm reduction over 73 years. Five of the 10 weather types had changes partly attributable to changing event frequency, which is partial attribution rather than a sole-cause estimate. Climate Dynamics, “Temporal trends and large-scale atmospheric and oceanic forcings of central U.S. snowfall producing weather systems, 1948–2021”
The study also describes substantial variability. Seasonal snowfall totals in the central United States had an average coefficient of variation of about 43%, while snowfall-event magnitudes had coefficients of variation above 70% at many locations; “many” is not a defined percentage of stations. Climate Dynamics, “Temporal trends and large-scale atmospheric and oceanic forcings of central U.S. snowfall producing weather systems, 1948–2021”
Regional comparisons show why geography and period must remain attached to every trend:
- Some locations in Minnesota, South Dakota, and Nebraska increased seasonal snowfall by as much as 0.5 cm per year from 1900 to 2009.
- In the same northern Plains region, a separate comparison found snowfall increases of more than 12 mm per year from 1948 to 2021.
- Moderate two-day snowstorms producing at least 15.2 cm became more frequent in parts of Minnesota, South Dakota, and Nebraska, although the source does not provide a percentage change.
- Seasonal snowfall in Missouri, Kansas, and parts of Iowa decreased by as much as 0.2% per year from 1930 to 2007.
- A separate 1948–2021 analysis found seasonal snowfall decreases of more than 2 mm per year in parts of Missouri, Kansas, and Iowa.
These findings use different periods, units, subregions, and weather-type groupings. They are regional observations and comparisons, not a single national estimate. Climate Dynamics, “Temporal trends and large-scale atmospheric and oceanic forcings of central U.S. snowfall producing weather systems, 1948–2021”
Winter Storm Damage, Fatalities, and Loss Statistics
NOAA’s billion-dollar disaster series counts events whose inflation-adjusted losses exceed $1 billion, so it excludes lower-cost winter storms. Damage figures are estimates and may be revised, while fatalities are event-summary counts.
The United States recorded two billion-dollar winter storm or cold-wave events in 2024. Both occurred across the Northwest and central or southern United States in mid-January, and the 2024 winter-weather disasters caused an estimated $3.6 billion in damage.
The two billion-dollar events drove the majority of that loss total, although the report does not quantify the share. NOAA, “2024: An active year of U.S. billion-dollar weather and climate disasters”
| Event geography and period | Estimated damage | Deaths |
|---|---|---|
| Central, southern, and northeastern United States, January 2024 | $1.9 billion | 41 |
| Pacific Northwest, January 2024 | $1.7 billion | 20 |
From 1980–2024, the United States recorded 24 winter storm events with losses exceeding $1 billion each. Across the thresholded severe-storm event set, the average cost of a billion-dollar severe-storm event was $2.5 billion; this is a NOAA-reported category average using CPI-adjusted costs, not an average for every winter storm. NOAA, “2024: An active year of U.S. billion-dollar weather and climate disasters”
Cold-Weather Mortality and Hypothermia Statistics
CDC’s 2023 figures use deaths attributed to excessive cold or hypothermia through underlying-cause ICD-10 code X31 and multiple-cause code T68. Death-certificate coding can miss cases, so these numbers describe coded mortality rather than every cold-related death.
The United States recorded 1,024 deaths in 2023 under that definition. The monthly distribution was:
| Month in 2023 | Share of coded annual deaths |
|---|---|
| January | 19.9% |
| February | 16.8% |
| March | 14.2% |
| April | 5.7% |
| May | 2.7% |
| June | 1.9% |
| July | 1.5% |
| August | 1.4% |
| September | 2.1% |
| October | 5.0% |
| November | 11.9% |
| December | 16.7% |
January had the largest monthly share at 19.9%, followed by February at 16.8%, March at 14.2%, and December at 16.7%. The distribution is by calendar month; it should not be treated as a complete definition of meteorological winter. CDC, “QuickStats: Percentage Distribution of Deaths Attributed to Excessive Cold or Hypothermia, by Month — United States, 2023”
Winter Weather Exposure and Vulnerable Population Data
A separate CDC analysis used a broader definition: any death during 2003–2013 with underlying or contributing ICD-10 X31 exposure to excessive natural cold. Under that definition, the United States recorded 13,419 hypothermia-related deaths, with annual rates ranging from 0.3 to 0.5 per 100,000 people.
Males accounted for 9,050 deaths, or 67%. Mean rates among people aged at least 65 were 1.8 per 100,000 for males and 1.1 per 100,000 for females.
Alcohol or drug poisoning was a contributing cause in 1,391 deaths, or 10%, which does not prove poisoning caused every death. CDC, “Hypothermia-Related Deaths — Wisconsin, 2014, and United States, 2003–2013”
Wisconsin surveillance during the record-cold period from January 1 through April 30, 2014 recorded 27 deaths. The small, event-specific investigation found:
- 18 of 27 decedents, or 67%, were male; the median age was 66 years, with a range of 25–95 years.
- 18 of 27 bodies, or 67%, were discovered outdoors, although discovery location is not necessarily exposure location.
- The median outside temperature at the estimated time of death was 6°F, with a range of −14°F to 35°F.
- Four decedents, or 15%, were found indoors in homes with unused or nonfunctional furnaces.
- Five decedents, or 19%, had a history of mental illness.
- Fifteen decedents, or 56%, lived alone, and two, or 7%, were homeless.
These Wisconsin percentages describe 27 investigated deaths during an unusually cold period. They should not be generalized to all winters or to the U.S. population. CDC, “Hypothermia-Related Deaths — Wisconsin, 2014, and United States, 2003–2013”
Winter Storm Forecasting and Measurement Context
Forecast products describe expected impacts and uncertainty rather than observed storm totals. NOAA’s deterministic Winter Storm Severity Index provides a three-day outlook of expected winter-storm impacts across the United States.
The probabilistic WSSI-P, operational from December 2023 and described in NOAA’s fiscal year 2024 report, provides a seven-day ensemble-based outlook. Its longer horizon and probabilistic design address uncertainty; neither product is a count of storms or a measurement of realized damage. NOAA, “Fiscal Year 2024 NOAA Annual Transition Report”
A broader climate index also needs careful labeling. NOAA reported that the 2023 U.S. Climate Extremes Index was 65% above average and ranked 11th highest in its 114-year record for the contiguous United States, but the index aggregates upper and lower 10% extremes in temperature, precipitation, drought, and landfalling tropical cyclones.
It is not a winter-only statistic. NOAA, “Assessing the U.S. Climate in 2023”