Typical Meteorological Year (TMY)

From Historical Trends to Future Plans

The Typical Meteorological Year (TMY) product provides a representative year of meteorological data built from decades of observations (1998–2022), as well as future projections through 2100 for specific locations across the United States. 

Developed with input from professionals across the Architecture and Engineering sector, this tool allows the industry to consider the risk of extreme weather events by incorporating applicable data into design scenarios to empower weather-resilient building and infrastructure designs.

Access TMY Data

Tell us how you use TMY data to help inform future improvements and updates. 

A single room cabin is buried halfway up the windows in snow; smoke rises from the chimney.
Source: Pexels.com, Mitchell Henderson

How do I use and download TMY data?

TMY Interactive Map

Use the TMY interactive map to create tailored calculations for different design scenarios based on TMY data and projections. This product is experimental; caveats on the use of this data can be found in the Technical User Guide. Contact industryproving.grounds@noaa.gov for more information.

Launch Interactive Map

Documentation

  • Technical User Guide: Caveats, calculation information, and other information on the TMY product.
  • User Guide Appendix: Example of how data is processed, list of GHCNh stations used and removed, list of POWER stations.

References

Update Schedule

  • NCEI’s Typical Meteorological Year is an experimental product. It leverages ISO standards but has not yet been peer-reviewed. Input data sources may be updated and cause slight variations to the output data. Please see the Technical User Guide or contact industryproving.grounds@noaa.gov for more information.
  • To provide feedback that informs the ongoing development of the TMY product and NCEI’s mission to deliver and disseminate information nationwide, select the button at the bottom right of the screen “Help improve this site.”

“Weather data and research on future weather extremes conducted by NOAA are vital to the work civil engineers do to protect the health, safety, and welfare of the public… Rainfall projections, temperature trends, and storm intensity data collected by NOAA inform the design of critical systems such as roads, bridges, stormwater networks, and coastal defenses.”

NCEI Architect and Engineering Working Group and ASCE Member

What is the TMY product?

TMY is an environmental data product designed to help users in understanding and analyzing historical conditions, and in adapting to both current and future projected meteorological conditions across the United States. The TMY product is designed to render high-resolution meteorological datasets via an accessible, interactive map interface.

What is a "typical" meteorological year?

TMY generates a year-long dataset composed of "typical" months selected by utilizing historical and modeled data. The historical product spans 1998-2022 and identifies the most typical months to represent each calendar month based on their statistical similarity to long-term weather norms. For example, if June 1998 is the most typical, it will be used for June in the TMY dataset. 

The result is a detailed, hourly dataset that represents a "typical" year for a given location that can facilitate weather-resilient design and planning throughout the United States. Future TMY values are determined by aligning historical analogs to future climate projections through 2100. For example, the future TMY dataset would adjust the air temperature in July 1999 for a given location based on the projected change in air temperature in July 2050 from 1998-2022 for that location.

How is TMY used?

TMY is especially useful for architects, engineers, and planners because it offers a practical and dependable resource to support designing infrastructure that is energy-efficient and resilient to current and future weather conditions in various climate zones across the United States.

TMY datasets can be used to: 

  • Simulate building energy demands.
  • Manage stormwater systems.
  • Understand annual variability in climatology for a specific location.
  • Inform models to evaluate the impact of future climate on infrastructure. 

Contact industryproving.grounds@noaa.gov for more information.

What variables are included in TMY?

Core Variables

Core variables captured in the TMY calculation include: 

  • Direct Normal Irradiance
  • Global Horizontal Irradiance
  • Dry-bulb Air Temperature
  • Dew Point Temperature
  • Wind Speed 

Additional Variables

Users can select the following additional variables: 

  • Wind Direction
  • Total Liquid Precipitation
  • Wind Gust
  • Relative Humidity
  • Snow Depth
  • Station Level Pressure
  • Total Sky Cover
  • Direct Horizontal Irradiance

Please note that snow depth and sky cover are not available for future projection output. The historical months that form the TMY are based on the core variables. Therefore, the additional variables are not guaranteed to be representative of typical conditions. Please see the Technical User Guide for more information.

What are the input data sources?

The following are the data sources used within the TMY methodology. See the Technical User Guide for more information on processing and access.

Product Details

The TMY product features include:

  • Interactive Map Interface: Users can easily access data for any location in the U.S., including Alaska, Hawaiʻi, Puerto Rico, and the U.S. Virgin Islands, through an intuitive point-and-click web map interface. This reduces technical barriers and makes the data widely accessible.
  • Typical Month Selection: Each month in the TMY dataset is chosen from historical data based on how closely it matches long-term average conditions. For example, if a user selects Miami, FL, the output shows February 2013 as the most typical February for the core variables.
  • Comprehensive Variables: TMY provides hourly data for multiple critical weather variables used in building and infrastructure design, such as temperature, solar radiation, humidity, and precipitation.
  • Current and Future Projections: This product includes a future projection component. This innovation integrates modeled environmental data (e.g., Geophysical Fluid Dynamics Laboratory SPEAR model projections) to simulate “typical” future conditions, enabling future-ready design.
  • Variable Weighting and Scenario Flexibility: Users will be able to tailor datasets based on different design needs or projection scenarios, providing adaptability for diverse applications in architecture, energy modeling, and infrastructure planning.
  • Downloadable Data: All data will be available for download in standard formats, suitable for use in simulation software and climate analysis tools.

User-Informed

NCEI engaged with professionals in the architecture and engineering at every stage of product development. A user group provided feedback on the interface design, functionality, and data output. The product was initiated based on listening sessions and workshops with members of the industry who needed this information as inputs to their building energy use simulations.

Such simulations inform infrastructure design and planning because they allow architects and engineers to examine the impact of annual variability in environmental conditions like irradiance, temperature, and wind speed, both historically and in the future.

Visit the Our Impact page to learn more about how NCEI’s products and services benefit U.S. industries. 

To provide feedback that informs the ongoing development of the TMY product and NCEI’s mission to deliver and disseminate information nationwide, please select the button at the bottom right of the screen “Help improve this site.”

Source Data

The historic portion of the product uses the following input data:

The future projections are from the NOAA Geophysical Fluid Dynamics Laboratory (GFDL) Seamless System for Prediction and EArth System Research (SPEAR) model.

How did we choose the stations?

The stations chosen for the TMY product come from the GHCN-hourly database. A two-phase filtering scheme is applied to determine stations that contain enough data for the TMY calculation. The first phase finds stations that contain at least 10 years of data across each of the 12 months. The second phase involves passing those stations through the TMY calculation to ensure the station would produce an adequate output dataset. For more details on the QA/QC process, see the Technical User Guide.

Historical Product

The TMY methodology for this product is based on the International Organization for Standardization (ISO) Standard 15927-4, with modifications from Wilcox and Marion (2008) and Beaufort et al. (2024).

The calculation begins by finding the daily means for all of the core parameters, daily maximums for air temperature, dew point temperature, and wind speed, and daily minimums for air temperature and dew point temperature. The cumulative distribution functions (CDFs) of the daily means, minimums, and maximums over all years in the data set for each month are determined. A separate CDF is found for the daily means, minimums, and maximums within each calendar month for each year.

Next, a Finkelstein-Schafer statistic is calculated for each month. A weighted sum is found from the Finkelstein-Schafer statistic and the weights applied on the climate parameters, which are user-specified. The individual years are ranked in order of increasing size of the weighted sum for each calendar month. Criteria are applied to remove years that have missing data. The most typical year for each month is chosen based on persistence criteria applied to ensure the data is representative of typical conditions, and does not include extremes or irregular values. The final hourly dataset are combined and smoothed across the month interface (last 8 hours and first 8 hours) to ensure continuity.

For more detailed information on the methodology, see the Technical User Guide.

Future Projections Product

The future projections portion of this product uses a morphing methodology based on Belcher et al. (2005) and Rodrigues et al. (2023). This approach adjusts the historic TMY data by the projected changes from a climate model.

This method uses air temperature, partial pressure of water vapor, wind speed, total precipitation rate, and relative humidity from the SPEAR climate model. An adjustment factor is calculated for each climate parameter between the early model time period (1998-2022) and the future model time period chosen by the user (2011-2035, 2026-2050, 2051-2075, or 2076-2100) under the SSP2-4.5 scenario. The adjustment factors are applied to the observed hourly TMY values to shift or stretch them. A similar smoothing process is done at the month interface to ensure continuity.

Only air temperature, dew point temperature, relative humidity, precipitation, and wind speed are morphed. Wind gust, wind direction, global horizontal irradiance (GHI), direct normal irradiance (DNI), and direct horizontal irradiance (DHI) are unchanged as this methodology is insufficient to accurately capture the projected changes of these parameters. Station-level pressure remains unchanged due to the expectation that this parameter will not change in the future. Sky cover and snow depth are not provided for future TMY, as the data was not available during this product’s development.

Three separate files are returned for future TMYs, containing the lower bound, mean, and upper bounds, respectively. The Technical User Guide contains detailed information on the calculation of uncertainty bounds for future TMY methodology used for this product. However, the methodology relies on a single model and one climate forcing scenario: the GFDL SPEAR Model for SSP2-4.5. The provided uncertainty reflects only that available from using one model and one scenario. This uncertainty may be larger than if multiple models and scenarios were used. Consequently, the future TMY should be viewed as just one potential future, not as a minimum, maximum, or expected outcome. For details on the underlying model uncertainty, please refer to the GFDL SPEAR page.

Weighting Scheme Information

It is important to select the weighting scheme that best aligns with the specific goals and requirements of your work. The following information is designed to assist you in choosing from the predefined weighting options; the custom weights options also allows users to tailor the weighting scheme as needed.

This product offers five different weighting schemes applied to the core variables in the TMY calculation for users to select from. Four schemes are predefined and based on existing schemes widely used for TMY products. The last is the Custom option, which allows for a user-specified weighting method. These weights are applied to the following ten variables: dry-bulb air temperature (minimum, maximum, and mean), dew point temperature (minimum, maximum, and mean), wind speed (maximum and mean), global horizontal irradiance, and direct normal irradiance. The following information outlines each predefined scheme:

This scheme places equal weights on the ten core variables used in the TMY calculation, with each weight set at 10%. The Standard weighting scheme is intended for use when core variables have an equal influence on the outcome. However, this method may not be suitable for all regional climates, where one meteorological variable may have a larger impact on building energy efficiency than another, such as solar radiation versus wind speed. This scheme is derived from ISO 15927-4.

This scheme places the largest weight on the mean global horizontal irradiance (at 50%) and no weight on the mean direct normal irradiance (0%). The minimum and maximum dry-bulb air temperature and dew point temperatures are set at 4.17%. The mean dry-bulb air temperature, mean dew point temperature, maximum wind speed, and mean wind speed are set at 8.33%. The Sandia weighting scheme is intended for use where the global horizontal irradiance will have the largest impact on building energy performance (Hall et. al., 1977).

This scheme places a weight of 25% on the mean global horizontal irradiance and the mean direct normal irradiance. The mean dry-bulb air temperature and mean dew point temperature are set at 10%. The rest of the variables are set at 5%. The NSRDB weighting scheme is intended for use in analyses relating to energy output. See the NSRDB TMY Data page for more information.

This scheme places the largest weights on the mean global horizontal irradiance (at 40%) and mean dry-bulb air temperature (at 30%). No weights are assigned to the direct normal irradiance variable (0%). The minimum dry-bulb air temperature, maximum dry-bulb air temperature, mean dew point temperature, maximum wind speed, and mean wind speed are set at 5%. The minimum and maximum dew point temperature are set at 2.5%. The Gaur and Lacasse weighting scheme is intended for use when the dry bulb temperature and global horizontal irradiance variables are more likely to affect building energy performance. See the paper by Gaur and Lacasse for more details.

To provide feedback that informs the ongoing development of the TMY product and NCEI’s mission to deliver and disseminate information nationwide, please select the button at the bottom right of the screen “Help improve this site.”

Contact industryproving.grounds@noaa.gov for more information.

To inform future updates and improvements, please tell us how you use TMY data by completing our optional User Registration Form.

The Typical Meteorological Year represents an hourly time series that best represents a location's long-term weather conditions. It is a concatenation of 12 typical meteorological months from statistically analyzed and selected individual months from the entire set of available years (see Technical User Guide documentation for more details). NCEI’s TMY product is delivered through a point-and-click interface with a downloadable dataset that provides one year’s worth of climatological data for the selected location chosen, displayed month-by-month from a record of many years of data.

This product consists of a historical product and a future projections product. The historical product spans 1998-2022 and identifies the most typical months to represent each calendar month. These months are chosen based on the core variables only.

The future projections product uses historical analogs that are adjusted based on climate model projections through 2100 to represent conditions that may be typical in the future. These adjustments may increase or decrease the value of the parameters in the future. For example, an adjustment of 3°C may be applied to the July 1999 temperatures to represent conditions predicted to occur in July 2050.

As the environment continues to change, a historically based “typical meteorological year” may become limited in usefulness. Due to the increasing frequency and intensity of extreme weather, historical data may become an unreliable basis for a TMY product intended for use in designing buildings that last for decades into the future.

The future projection component of this product aims to close this gap by providing one example case of a potential future. Note that this product is a single estimate of how typical weather conditions may be represented in the future and is not a guarantee of future weather that may occur. This dataset is most useful alongside additional resources and data to gain a better understanding of what future conditions may look like.

This product spans the contiguous U.S., Alaska, Hawaiʻi, Puerto Rico, and the U.S. Virgin Islands. There are 1,972 GHCNh stations included in this product.

The historical component includes data from 1998 to 2022. The future projections component uses projections through 2100, which are binned in the following periods: 

  • 2011-2035
  • 2026-2050
  • 2051-2075
  • 2076-2100

The output file is available as a CSV file with 8,760 rows and 8 to 13 columns of variables and an EnergyPlus Weather (EPW) file with 8,760 rows and 35 columns.

For the future projections output, two accompanying files show the lower and upper bounds of the data.

The output files contain hourly data for a representative typical year. Because the product selects the 'most typical' month over its historic dataset to represent the output, the hourly data may come from different years. For example, the data may be from January 2005, February 2016, and so on.

The core variables for the calculation include: direct normal irradiance, global horizontal irradiance, dry-bulb air temperature, dew point temperature, and wind speed.

Users can select the following additional variables: wind direction, total liquid precipitation, wind gust, relative humidity, snow depth, station level pressure, total sky cover, and direct horizontal irradiance.

Note that for the future projections, snow depth and sky cover are not available.

For the historical component, data comes from the National Laboratory of the Rockies’s National Solar Radiation Database(NSRDB) and the NCEI’s Global Historical Climatology Network hourly (GHCNh) dataset.

As a supplement to the National Solar Radiation Database for stations poleward of 60°N, data comes from the NASA Prediction Of Worldwide Energy Resources (POWER) version 2.4.9.

For the future projections component, the NOAA’s Geophysical Fluid Dynamics Laboratory Seamless System for Prediction and Earth System Research model for the SSP2-4.5 scenario is being used.

For more information on these datasets and the input data for this product, see the Technical User Guide.

Updates to the TMY product are not planned at this time. To inform future updates and improvements, please tell us how you use TMY data by completing our optional User Registration Form. Contact industryproving.grounds@noaa.gov for more information.

Updates to the TMY product, including additional projection models and climate forcing scenarios, are not planned at this time. This product is using the GFDL SPEAR model for the SSP2-4.5 scenario. For more information, see the Technical User Guide.

This dataset is experimental and has not yet been peer-reviewed. The calculation methodology explained in the Technical User Guide was derived from various peer-reviewed sources.

It is anticipated that this product will be applicable for use in building energy simulation software, such as EnergyPlus. When used for engineering or architectural purposes, it is advised to continue using trusted resources alongside this product.

To inform future updates and improvements, please tell us how you use TMY data by completing our optional User Registration Form.

There is no guarantee that the Additional Variables will illustrate a "typical" amount for the year selected. The TMY calculation does not take into account the Additional Variables, only the Core Variables. Instead, the Additional Variable data will be pulled from the same year as the Core Variables for that month. For example, the associated precipitation values for a location may be taken from a year that had more rainfall than typical for that year because the TMY calculation is based on core variables of solar radiation, temperature, and wind speed.

For some rare stations, the data isn't sufficient to use for the TMY calculation. In this case, the second or third nearest stations were used in place of the absolute nearest station.

NCEI can certify data that has been archived. This product is not archived, and therefore is not certifiable. See the NCEI Data Certification page for more information.

This error could indicate issues on NCEI’s end. If this issue persists, please try again later. Reach out to industryproving.grounds@noaa.gov if this persists.