A Primer on Atlas 15

This primer is excerpted from Mid-Atlantic Regional Climate Impacts Summary and Outlook: Winter 2025–2016.

Introduction

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.¹ 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.^2,3 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.⁴ 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 14.⁵ 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.⁶

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 1 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 1 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.