Mid-Atlantic Regional Climate Impacts Summary and Outlook: Winter 2025–2026

Highlights

• Winter temperatures were below normal for the entire region, with most locations seeing temperatures that were 2–4 degrees below normal.
• Most of the region was drier than normal, with drought covering much of Maryland and Virginia throughout the winter season.
• Several snowstorms impacted the region, causing power outages, school closures, and vehicle crashes. Prolonged stretches of cold temperatures allowed snow to stay on the ground much longer than normal in several locations in Maryland and Virginia.
• A primer on NOAA precipitation frequency estimates, particularly Atlas 14 and Atlas 15, is included in this summary in anticipation of the Atlas 15 data release later this year.

View all climate summaries

This summary focuses on winter weather and climate events in the Chesapeake Bay watershed and provides highlights from the greater Mid-Atlantic region. The winter season is defined as the months of December, January, and February. The MARISA region covers Maryland, Delaware, Virginia, and Pennsylvania and the portions of New York and West Virginia that fall within the boundaries of the Chesapeake Bay watershed, as shown in Figure 1 below. We refer to this region as the Mid-Atlantic region in the rest of the climate summary.

This map shows the “MARISA region”. The lightly shaded area shows the extent of the Chesapeake Bay Watershed.

Part 1: Significant Weather Events and Impacts

Winter Weather

On December 8, a storm brought snow to Virginia, with the greatest totals ranging from 5 to 7 inches.¹ High snowfall rates of 1 to 2 inches per hour were reported in parts of central and eastern Virginia, caused by a passing front.² For several sites, it was the most snow seen on a December day in years.³ For example, Richmond accumulated 4.0 inches of snow, which made December 8 its snowiest December day since 2018, while Norfolk measured 0.8 inches, its snowiest December day since 2010.⁴ The storm resulted in hundreds of vehicle crashes across the state.⁵

A storm from December 26 to 27 brought a mix of precipitation types to the watershed⁶, which disrupted travel after the Christmas holiday.⁷ The greatest snowfall totals of 6 to 12 inches were reported in central New York.⁸ Ice accumulations of 0.10 to 0.40 inches were noted in places like central Pennsylvania and western and central Maryland.^9,10 The weather caused flight delays and cancellations and hazardous road conditions.^11,12

Snow squalls moved through the watershed from late on December 31 into the morning of January 1, creating hazardous travel conditions.^13,14 For instance, reports from Pennsylvania noted wind gusts of 30 to 50 miles per hour (mph), near-zero visibility, and snow-covered roadways.¹⁵

A winter storm brought snow, ice, and rain to the watershed from January 24 to 26. The greatest storm snowfall totals of 12 to 18 inches were reported in central Pennsylvania and central New York.^16,17 January 25 became one of the 10 snowiest January days on record for several sites including Harrisburg and Scranton, Pennsylvania; Dulles Airport, Virginia; Baltimore, Maryland; and Binghamton, New York.¹⁸

Streets in downtown Washington, D.C. remained snow-packed following a storm that paralyzed transportation, closed government, and shut down schools, January 26, 2026.

Photo by Andrew Leyden/ZUMA Press Wire/Reuters

Portions of Maryland, West Virginia, and Virginia also saw sleet and/or freezing rain.¹⁹ Sleet accumulated several inches deep in some areas, while ice accretion from freezing rain was 0.25 inches or less in most locations.^20,21,22 During the storm, several Maryland and Virginia airports including BWI, Reagan National, Dulles, and Richmond, reported more than 10 hours of sleet, ranking among the 10 greatest number of such hours for January at each site.²³ In fact, Baltimore experienced a total of 16 hours of sleet during January (11 during this storm and five during a mid-month storm), its greatest number of hours of sleet for January and tying its second greatest for any month, with records going back to 1949.²⁴ The combination of sleet with cold temperatures (see discussion below) created thick layers of ice that were so heavy that it broke snowplows or required construction equipment to remove.²⁵

Travel was difficult due to flight delays and cancellations, limited rail and bus service, and hazardous road conditions.²⁶ For example, on January 25, all flights at Reagan National Airport and Harrisburg International Airport were cancelled.^27,28 Other storm impacts included scattered power outages and school closures.^29,30 Some schools in the Washington, D.C. metropolitan area were closed for over a week due to challenges in clearing icy roads and sidewalks.³¹

Source: NOAA National Operational Hydrologic Remote Sensing Center Interactive Snow Information tool, 2026

From February 22 to 23, a storm brought heavy snow and high winds to coastal parts of the Mid-Atlantic including the Delmarva Peninsula.^32,33 Snowfall totals were up to 21 inches in central and southern Delaware, while parts of Maryland's Eastern Shore and Virginia's Accomack County saw between 8 and 14 inches of heavy, wet, snow, falling at a high rate of up to 3 inches per hour.^34,35 Wind gusts of up to 55 miles per hour were recorded, briefly causing blizzard-like conditions.³⁶ The combination of heavy, wet snow and winds led to downed trees and power outages.^37,38 In some locations, power restoration took several days.³⁹ For much of the rest of the Chesapeake Bay watershed, snowfall totals were 8 inches or less.⁴⁰

Source: NOAA National Operational Hydrologic Remote Sensing Center Interactive Snow Information tool, 2026

Source: https://www.weather.gov/akq/Feb22232026_WinterStormSummary

Extreme Cold

There were a few cold air outbreaks during January, with temperatures well below normal during one such event running from January 23 to February 1.⁴¹ Some of the coldest temperatures were reported in central Pennsylvania and central New York, where, at times, lows were around -15 degrees F and highs were in the single digits.⁴² Subzero wind chills were observed across much of the watershed.⁴³ The number of consecutive days with high and/or low temperatures meeting certain thresholds was noteworthy at multiple sites.⁴⁴ For example, Scranton, Pennsylvania reported nine consecutive days with a high of less than 20 degrees F, its third longest such streak on record.⁴⁵ Similarly, Harrisburg, Pennsylvania and Dulles Airport, Virginia experienced their second-longest streak of days with a high less than 25 degrees F at eight days and four days, respectively.⁴⁶ Due to these extremely cold temperatures, Dulles Airport went on to tie its greatest number of January days with a high less than 25 degrees F at six days.⁴⁷ The cold weather contributed to numerous water main breaks, caused other pipes to freeze, and made snow and ice removal difficult.^48,49,50 There was an uptick in emergency room and urgent care visits in Maryland and several deaths were reported in Virginia that were tied to the cold temperatures.^51,52

Unusually cold temperatures persisted through the first 10 days of February, particularly from February 5 to 9.^53,54 Gusty winds led to subzero wind chills in many areas, with the lowest around -30 degrees F in central New York and northern Pennsylvania.^55,56 The cold weather contributed to several deaths in Maryland.⁵⁷ The prolonged stretch of cold weather from late January through early February allowed ice to build up in the Chesapeake Bay, reaching 38 percent coverage as of February 9.⁵⁸ Some boats became iced in, temporarily halting oyster harvesting operations and cancelling some charter services.⁵⁹

Cold temperatures helped keep snow on the ground for an extended period from late January through mid-February.⁶⁰ Dulles Airport, Virginia and Baltimore, Maryland had 18 consecutive days (January 25 to February 11) with four or more inches of snow depth.⁶¹ This tied the longest such streak at Dulles and was the second-longest such streak for Baltimore.⁶²

Drought

The U.S. Drought Monitor from December 2 showed most of the watershed in drought or abnormal dryness.⁶³ Severe drought was found in eastern West Viriginia and northern Virginia, while moderate drought covered parts of Virginia, Maryland, eastern West Virginia, central Pennsylvania, and central New York.⁶⁴ In fact, drought covered over half of Virginia, a third of Maryland, and 19 percent of Pennsylvania.⁶⁵ Abnormal dryness was present in much of the rest of the watershed.⁶⁶ Below-normal precipitation in December allowed conditions to intensify in many areas over the course of the month.⁶⁷ Severe and moderate drought expanded in Virginia, easter West Virginia, Maryland, and Pennsylvania.⁶⁸ The U.S. Drought Monitor from January 6 showed drought covering 74 percent of Maryland, 72 percent of Virginia, and 45 percent of Pennsylvania.⁶⁹

Groundwater levels were below normal or lower for many parts of the watershed during the month of December.⁷⁰ Record low groundwater levels were measured at times in locations such as central Pennsylvania and central and southern Maryland.⁷¹ Water conservation continued to be encouraged in the region, with a few water suppliers having mandatory water restrictions.⁷²

Drought conditions deteriorated in some areas but improved in others during the month of January.⁷³ For example, severe and/or moderate drought spread in southern Pennsylvania and central and eastern Virginia but contracted in northern Pennsylvania and western Virginia.⁷⁴ The U.S. Drought Monitor from January 27 showed drought covering 74 percent of Maryland, 76 percent of Virginia, and 36 percent of Pennsylvania.⁷⁵

During January, record to near-record low streamflow, groundwater levels, and soil moisture were observed at times in locations such as southern Pennsylvania, central Maryland, and eastern West Virginia.^76,77 Mandatory water restrictions were in place for a few locations including Purcellville and Woodstock, Virginia, and parts of three central Pennsylvania counties.^78,79,80 Water conservation was encouraged in many other places across the watershed.^81,82 Water levels on Raystown Lake in southern Pennsylvania declined due to the drought conditions, leading to the closure of a boat ramp.⁸³

During February, conditions deteriorated in northern parts of the watershed but improved in the southern areas.⁸⁴ For example, due to below-normal precipitation and declining streamflow and groundwater levels, moderate drought spread in central New York and northern and central Pennsylvania.⁸⁵ However, beneficial precipitation and improving streamflow and groundwater levels allowed severe drought to be erased from eastern West Virginia and significantly contract in Maryland and Virginia.⁸⁶ The U.S. Drought Monitor from February 24 showed severe drought covered 17 percent of Virginia and only two percent of Maryland compared to 44 and 43 percent, respectively, as of January 27.⁸⁷ However, Pennsylvania's drought coverage grew to 50 percent as of February 24 compared to 36 percent at the end of January.⁸⁸

Reduced streamflow and groundwater levels persisted in parts of the watershed during February.^89,90 Below-normal soil moisture was also found in some areas.⁹¹ Mandatory water restrictions continued or were implemented in a few locations, particularly central Pennsylvania.^92,93 Water conservation continued to be encouraged.⁹⁴

Drought persisted in many areas this winter despite several snowstorms due to several factors. Generally, high pressure systems dominated the weather pattern this winter, which limited the number of storms and overall precipitation totals.⁹⁵ These few storms produced widely variable precipitation amounts and could not compensate for the overall lack of precipitation during the winter season and longer-term deficits present in some locations.⁹⁶ Additionally, reduced groundwater levels continued to be a driver of drought in some locations, as groundwater recovery is slower than surface water recovery, and in areas where the ground is frozen, recovery will not happen until the spring thaw.⁹⁷

SOURCE: U.S. Drought Monitor

Part 2: Seasonal Temperature and Precipitation

Temperature

Figure 7 shows the winter 2025-2026 average temperature compared with the climate normal—i.e., the average seasonal temperature from 1991 to 2020.⁹⁸ The figure shows that nearly the entire Mid-Atlantic region experienced below normal temperatures (from 2-6 degrees below normal) this winter. The region also experienced colder than normal temperatures last winter (winter 2024-2025), when most of the region saw temperatures between 0 and 4 degrees below normal.

SOURCE: Northeast Regional Climate Center, 2025 (https://www.nrcc.cornell.edu). Used with permission.

Normal temperature is based on the winter season's average temperature data from 1991–2020. Shades of red indicate above-normal temperatures. Shades of blue indicate below-normal temperatures. The boundaries of the Chesapeake Bay watershed are outlined in bold black. Average departure from normal temperature is based on comparing the Parameter Elevation Relationships on Independent Slopes Model (PRISM) gridded dataset from the fall season to the PRISM gridded historical dataset from 1991-2020. Additional details about these data are available at https://www.rcc-acis.org/docs_gridded.html.

Winter 2025-2026 was among the 20 coldest winters on record for two sites (Table 1). It was the coldest winter since winter 2014-2015 for several sites in the watershed.^99,100

Station Name	Avg. Temp (degrees F)	Normal Temp (degrees F)	Rank (coldest)
Salisbury, MD	34.5	38.7	14
Binghamton, NY	22.2	25.0	20

SOURCE: Northeast Regional Climate Center, 2026 (https://www.nrcc.cornell.edu). Used with permission.

Monthly Temperature Rankings

Table 2 shows that this December was among the 20 coldest Decembers on record for Binghamton, New York and Dulles Airport, Virgina. It was also the coldest December since 2010 for multiple sites across the watershed.¹⁰¹ Despite the extreme cold temperatures described in Part 1 in January, only one site experienced temperatures over the course of the month that ranked among their top 20 coldest.

The full set of monthly rankings, locations, and temperatures are shown in Table 2.

December Temperature Rankings (coldest)
Station Name	Avg. Temp (degrees F)	Normal Temp (degrees F)	Rank (coldest)
Binghamton, NY	23.4	28.1	13
Dulles Airport, VA	34.6	37.7	20

January Temperature Rankings (coldest)
Station Name	Avg. Temp (degrees F)	Normal Temp (degrees F)	Rank (coldest)
Charlottesville, VA	33.1	38.4	20

February Temperature Rankings (coldest)
Station Name	Avg. Temp (degrees F)	Normal Temp (degrees F)	Rank (coldest)
No sites experienced temperatures that ranked in their top 20 coldest or warmest months of February on record.

SOURCE: Northeast Regional Climate Center, 2026 (https://www.nrcc.cornell.edu). Used with permission.

Precipitation

Figure 8 shows how the total precipitation for December 1, 2025, through February 28, 2026, differed from normal, with normal being defined as the average winter precipitation from 1991–2020. This figure shows that nearly the entire region experienced less than normal amounts of precipitation. The northern and southern portions of the region experienced between 75 and 100 percent of normal precipitation while the central portion of the region was a little drier, experiencing between 50 and 75 percent of normal precipitation. A few areas in far western Pennsylvania and West Virginia experienced even drier conditions, receiving between 25 and 50 percent of normal precipitation.

SOURCE: Northeast Regional Climate Center, 2026 (http://www.nrcc.cornell.edu). Used with permission.

NOTE: Normal seasonal precipitation is based on precipitation data from 1991–2020. Brown shades indicate below normal seasonal precipitation. Blue shades indicate above normal seasonal precipitation. The boundaries of the Chesapeake Bay watershed are outlined in bold black. Average departures from normal precipitation are based on the Northeast Regional Climate Center's interpolated gridded precipitation data for the winter season compared with the National Centers for Environmental Information's gridded normals for the winter season, which is based on historical data from 1991-2020. Additional information about the Northeast Regional Climate Center's gridded datasets can be found at https://www.rcc-acis.org/docs_gridded.html.

Winter 2025-2026 ranked among the 20 driest winters on record for five sites across the watershed (Table 3).

Station Name	Precipitation (inches)	Normal Precipitation (inches)	Rank (driest)
Baltimore, MD	6.06	9.69	11
Williamsport, PA	5.24	8.54	13
Salisbury, MD	7.38	10.35	14
Harrisburg, PA	6.05	9.05	16
Scranton, PA	4.88	7.46	19

SOURCE: Northeast Regional Climate Center, 2026 (https://www.nrcc.cornell.edu). Used with permission.

Monthly Precipitation Rankings

December was among the 20 driest months of December on record for two sites, Dulles Airport, Virginia and Baltimore, Maryland. Dulles Airport, Virginia, then experienced its 19^th wettest January on record. This February was the second-driest February on record for Scranton, Pennsylvania, and the 14^th driest for Williamsport, Pennsylvania.

The full set of monthly rankings, locations, and amounts of precipitation are shown in Table 4.

December Precipitation Rankings (driest)
Station Name	Precipitation (inches)	Normal Precipitation (inches)	Rank (driest)
Dulles Airport, VA	1.30	3.30	10
Baltimore, MD	1.46	3.71	20

January Precipitation Rankings (wettest)
Station Name	Precipitation (inches)	Normal Precipitation (inches)	Rank (wettest)
Dulles Airport, VA	3.18	2.94	19

February Precipitation Rankings (driest)
Station Name	Precipitation (inches)	Normal Precipitation (inches)	Rank (driest)
Scranton, PA	0.63	2.07	2
Williamsport, PA	1.04	2.31	14

Source: Northeast Regional Climate Center, 2026 (https://www.nrcc.cornell.edu). Used with permission.

Snowfall

Figure 9 shows how the total snowfall for December 1, 2025, through February 28, 2026, differed from normal, with normal being defined as the average winter snowfall from 1991 – 2020. The amount of snowfall compared to normal varied widely across the region. Portions of southern Virginia, easternmost Pennsylvania, and the Delmarva Peninsula saw over 200 percent of normal snowfall. While a central portion of the region covering portions of central Pennsylvania, western Maryland, northern Virginia, and eastern West Virginia saw 75 to 50 percent of normal amounts of snowfall. Most of the rest of the region experienced between 125 and 75 percent of normal snowfall.

SOURCE: Northeast Regional Climate Center, 2026 (http://www.nrcc.cornell.edu). Used with permission.

NOTE: Normal seasonal snowfall is based on snowfall data from 1991–2020. Oranges indicate below-average seasonal snowfall, and purples indicate above-normal seasonal snowfall. Percentages of normal seasonal snowfall are based on station-specific normal seasonal snowfall for the winter compared to the same station's winter 2025–2026 seasonal snowfall. Station-level percentages of normal are then spatially interpolated to form the figure above.

Salisbury, Maryland had its 16^th snowiest winter on record (Table 5).

Station Name	Snowfall (inches)	Normal Snowfall (inches)	Rank (snowiest)
Salisbury, MD	14.8	6.8	16

Source: Northeast Regional Climate Center, 2025 (https://www.nrcc.cornell.edu). Used with permission.

This December was among the top 20 snowiest on record for four sites and January was among the top 20 snowiest on record for one site. Scranton, Pennsylvania was the only site to experience top 20 snowiest months for both months. Scranton followed this up with a top 20 least snowy February.

The full set of monthly rankings, locations, and amounts of snowfall are shown in Table 6.

December Snowfall Rankings (snowiest)
Station Name	Snowfall (inches)	Normal Snowfall (inches)	Rank (snowiest)
Binghamton, NY	27.9	18.1	11
Richmond, VA	6.0	1.8	16
Lynchburg, VA	6.6	2.0	17
Scranton, PA	14.3	7.7	19

January Snowfall Rankings (snowiest)
Station Name	Snowfall (inches)	Normal Snowfall (inches)	Rank (snowiest)
Scranton, PA	17.3	11.7	19

February Snowfall Rankings (snowiest)
Station Name	Snowfall (inches)	Normal Snowfall (inches)	Rank (snowiest)
Salisbury, MD	10.8	3.4	9

February Snowfall Rankings (least snowy)
Station Name	Snowfall (inches)	Normal Snowfall (inches)	Rank (least snowy)
Williamsport, PA	2.6	9.3	14
Scranton, PA	3.6	10.9	19

Source: Northeast Regional Climate Center, 2026 (https://www.nrcc.cornell.edu). Used with permission.

Part 3: Spring 2026 Outlook

Temperature and Precipitation

As of February 19, 2026, the NOAA Climate Prediction Center forecasts an equal chance of below, near-, or above normal temperatures for much of the region for the months of March, April, and May 2026.¹⁰² Southern West Virginia, much of Virginia, and southern Maryland have a 33-40-percent chance of above normal temperatures for the spring season.¹⁰³ The precipitation forecast shows that the majority of the region is forecasted to have an equal chance of near-normal, above normal, and below normal precipitation for March, April, and May, while western New York and western Pennsylvania are forecasted to have a 33-40-percent chance of above normal precipitation.¹⁰⁴

Drought Incidence

The U.S. Seasonal Drought Outlook identifies how drought might change across the United States and categorizes areas by whether drought could develop or become more or less intense. As of February 28, 2026, the Outlook shows that drought is expected to persist across much of the region, including nearly all of Virginia and Maryland, and central and eastern Pennsylvania in the March 1 through May 31, 2026, timeframe.¹⁰⁵ Drought is expected to be removed in a small area of western Pennsylvania and in some of western New York.¹⁰⁶ Elsewhere, no new areas of drought are anticipated.¹⁰⁷

Climate Circulation Patterns

NOAA's Climate Prediction Center, which monitors the likelihood of occurrence of El Niño and La Niña climate phenomena, has a La Niña advisory active as of February 12, 2026.¹⁰⁸ There is a 60% chance that conditions will transition to ENSO-neutral in the February – April 2026 timeframe.¹⁰⁹ ENSO-neutral means that neither La Niña nor El Niño conditions are present, and as a result, "global seasonal conditions [will be] less predictable".¹¹⁰

ENSO conditions are one of the factors taken into account in NOAA's long-term forecasts and seasonal outlooks such as the one included in this climate summary.¹¹¹ However, other regional climate dynamics and natural climate variability also influence weather in the Mid-Atlantic. Additional information on La Niña and El Niño is available from the Pacific Marine Environmental Laboratory (La Niña, El Niño).

Part 4: A Primer on Atlas 15

Understanding precipitation frequency estimates is fundamental to hydrologic and infrastructure design. Engineers and planners use these estimates (e.g., the "100-year storm") to establish criteria for designing and managing stormwater infrastructure, such as detention basins, culverts and drainage systems. Beyond stormwater, precipitation frequency data also inform floodplain mapping and risk assessment, transportation drainage design, water resources planning such as reservoir and spillway sizing, and regulatory compliance.

Precipitation frequency analysis quantifies the statistical likelihood that rainfall of a given magnitude and duration will occur. Results are typically expressed through Intensity‑Duration‑Frequency (IDF) or Depth‑Duration‑Frequency (DDF) curves. Within an IDF or DDF curve, each estimate represents the expected rainfall depth associated with a defined exceedance probability, for example, a 1 percent annual exceedance event, commonly termed the "100‑year storm." The reliability of these estimates depends directly on the quality of observational data and the suitability of the statistical methods applied.

The primary source of precipitation‑frequency estimates in the United States has been NOAA's Atlas 14 database, which dates back to 2004. NOAA plans to release a major update to Atlas 14—Atlas 15, constituting an advancement in U.S. precipitation frequency estimation. As Atlas 15 begins to replace its predecessor, it also introduces important methodological changes in how extreme rainfall is characterized and applied in infrastructure and hazard analyses. This primer provides a background on Atlas 14 and outlines key changes anticipated in Atlas 15.

Atlas 14

NOAA's Atlas 14, first released in 2004 and updated through 2021, replaced older regional precipitation frequency atlases with a unified set of gridded datasets derived from rain gauge networks.¹¹² The Atlas 14 data and accompanying documentation are publicly available through NOAA's Precipitation Frequency Data Server (PFDS). Since its release, Atlas 14 has become the national standard for precipitation frequency information, widely incorporated into engineering design, planning and policy frameworks across the United States.

In recent years, observed precipitation trends have shown notable increases in both the frequency and intensity of short‑duration, high‑intensity events—especially in the Mid‑Atlantic and Northeast regions.^113,114 Atlas 14, however, is based on a stationary statistical framework, which assumes that the underlying climate and distribution of extreme rainfall remain constant over time.¹¹⁵ This assumption is increasingly inconsistent with observed and projected climate trends, underscoring the need for updated methods and data, such as those anticipated in Atlas 15.

The Atlas 14 dataset is also limited in spatial continuity and temporal coverage:

• Each regional volume was analyzed independently, creating subtle discontinuities at interstate boundaries (e.g., between Pennsylvania and Maryland).
• For much of the Mid‑Atlantic, the gauge record incorporated into Atlas 14 ends around 2006, omitting nearly two decades of subsequent high‑intensity convective and tropical rainfall events.
• Limited sub‑hourly data were available; most estimates relied on hourly or daily observations, constraining the accuracy of very short‑duration design storms.

Atlas 15

NOAA's Atlas 15 is being developed by the Office of Water Prediction to update and refine Atlas 14.¹¹⁶ The project responds to the known limitations in Atlas 14, including its stationary analytical framework, inconsistencies between regional volumes, and truncated periods of record in many parts of the country. Funding from the Bipartisan Infrastructure Law (BIL) supports this effort as part of a broader initiative to modernize federal climate and hydrology datasets used in infrastructure and hazard planning.¹¹⁷

The development of Atlas 15 reflects three main areas of advancement.

1. The first is spatial continuity. The new Atlas replaces the region‑by‑region design of Atlas 14 with a single, nationally consistent dataset that uses spatially continuous statistical models. This structure removes artificial boundaries between states and watersheds, producing smoother transitions in precipitation‑frequency values across the United States and territories.
2. The second is temporal dynamics. Atlas 15 introduces non‑stationary statistical methods that allow the parameters of the precipitation‑frequency distributions to vary through time. This approach recognizes that the frequency and magnitude of extreme rainfall may change, enabling the development of time‑specific (rather than fixed historical) return‑period estimates.
3. The third is climate integration. For the first time, Atlas 15 incorporates information from downscaled global climate models, such as the CMIP6 ensemble. These data are used to derive adjustment factors that describe how precipitation intensity and frequency could evolve over the remainder of the century.

Atlas 15 is organized into two volumes, as shown in Figure 10 below, that together provide a link between observed extremes and projected future conditions.

• Volume 1 will present updated precipitation‑frequency estimates derived from the extended historical record, expected to cover roughly 1950 through the early 2020s. This volume supersedes Atlas 14 as the best estimate of historical climate precipitation frequency estimates.
• Volume 2 will provide model‑based adjustment (change) factors that can be applied to the Volume 1 data to evaluate potential future changes in extreme rainfall, extending coverage through about the year 2100.

SOURCE: ASCE, 2025: https://collaborate.asce.org/blogs/committee-on-adaptation-to-a-changing-climate/2024/09/12/predicting-future-precipitation-intensity-duration

The figure titled “NOAA Atlas 15” shows the new national precipitation frequency standard, comparing historical and projected precipitation frequency estimates.

The chart displays precipitation intensity (in inches) on the vertical axis from 4 to 10 inches, and time on the horizontal axis from the year 1900 to 2100. The figure compares historical stationary and non-stationary intensity-duration-frequency (IDF) estimates for 1-day rainfall events of 100-year and 2-year return periods.

• Blue solid line: Stationary Historical IDFs (NOAA Atlas 14), showing a relatively stable trend over time.
• Blue dotted line: Non-stationary Historical IDFs (NOAA Atlas 15), showing a moderate increase beginning around the year 2000.
• Orange lines and shaded region: Future IDF projections under climate-adjusted scenarios, showing a wider range of precipitation values and increasing intensity through the year 2100.

The figure includes two labeled sections along a red double-headed arrow at the bottom:

• Volume 1: Based on historical gages and observed trends (covering roughly 1900–2020).
• Volume 2: Incorporates climate projection adjustment factors (covering roughly 2020–2100).

Table 6 summarizes the principal methodological and data structure differences between Atlas 14 and 15.

Category	Atlas 14	Atlas 15
Statistical Framework	Stationary GEV using regional L‑moment estimation; fixed climate assumption.	Non‑stationary GEV with hierarchical Bayesian estimation; parameters vary with time and covariates.
Spatial Modeling	Discrete regional zones analyzed independently; boundary discontinuities.	Spatially continuous geostatistical modeling with covariate‑driven smoothing.
Data Record	Gauge data largely ending ~2006 for much of the Mid‑Atlantic; limited sub‑hourly resolution.	Extended record through ~2022, integrating automated gauges, mesoscale networks, and radar‑derived precipitation fields.
Uncertainty Treatment	Regional confidence intervals estimated independently.	Spatial‑temporal uncertainty derived from Bayesian posteriors, enabling probabilistic mapping.
Temporal Products	Single historical climatology.	Two volumes—Volume 1: updated baseline; Volume 2: future projections via climate‑model‑based adjustments.
Data Access	Regional PFDS web portal with tabulated IDF data.	NOAA Atlas 15 portal with interactive visualization, time‑conditioned queries, and both historical and future frequencies.

Practical Implications

Atlas 15 represents an important update to the United States' precipitation frequency dataset—moving from a stationary, region by region framework to a dynamic, climate change informed, and spatially continuous model. Engineers can expect discernible changes in precipitation frequency values within Volume 1, particularly where intensification of short-duration events has been observed in recent years. Volume 2's projection-based adjustment factors will provide an explicit consideration of future changes in extreme rainfall within design analyses. For the Mid-Atlantic, the inclusion of post-2006 precipitation records, higher-resolution gauge and radar data, and integration of climate projections will produce more spatially consistent and climate relevant precipitation frequency estimates, potentially increasing future design storm intensities.

Timeline and Availability

Atlas 15 development follows a phased schedule. Following the release of a pilot dataset for Montana in late 2024, preliminary data for the contiguous United States (CONUS) will be subjected to public peer review in early 2026, with publication and adoption anticipated later that year. Datasets for areas outside the contiguous United States (OCONUS)—including Alaska, Hawaii, Puerto Rico, and Pacific territories—are planned for peer review in 2026 and publication in 2027.

Once complete, Atlas 15 will formally supersede Atlas 14 as the national standard for precipitation frequency estimation.