AKALARAZCACOCTDEFLGAHIIAIDILINKSKYLAMAMDMEMIMNMOMSMTNCNDNENHNJNMNVNYOHOKORPARISCSDTNTXUTVAVTWAWIWVWYDCUSAPRWPI Skip to main content

NOAA National Centers
for Environmental Information

State Climate Summaries 2022

TECHNICAL DETAILS & ADDITIONAL INFORMATION

Alaska Glacier
Image by Jacqueline Schmid from Pixabay

1. Overview

The information given here describes technical details and additional information related to the 2022 State Climate Summaries, including the most common datasets and analyses used. For elements not covered here, please refer to the metadata that accompany each figure.

a. Historical Climate Conditions

The description of historical climate conditions for each state is based on an analysis of core climate data (the primary data sources are described below). However, to help understand, prioritize, and describe the importance and significance of different climate conditions, additional input was derived from climate experts in each state, some of whom are authors on these state summaries. In particular, input was sought from the NOAA Regional Climate Centers and from the State Climatologists. The historical climate conditions are meant to provide a perspective on what has been happening in each state and what types of extreme events have historically been noteworthy, to provide a context for assessment of future impacts.

b. Future Climate Scenarios

The future climate scenarios are intended to provide an internally consistent set of climate conditions that can inform analyses of potential impacts of climate change. The scenarios are not intended as projections, as there are no probabilities for their future realization attached. They simply represent an internally consistent climate picture under certain assumptions about the future pathway of greenhouse gas emissions. By “consistent” we mean that the relationships among different climate variables and the spatial patterns of these variables derive directly from the same set of climate model simulations and are therefore physically plausible. The future climate scenarios are based on well-established sources of information. No new climate model simulations or downscaled datasets were produced for use in these state summaries.


2. Datasets

The most common datasets used in the State Climate Summaries are described here. Information on additional datasets used can be found in the metadata that accompany each relevant figure.

a. Past Climate

Historical seasonal and annual temperature and precipitation conditions for the contiguous United States and Alaska were analyzed using data from the NOAA NCEI’s Climate Divisional Dataset (nClimDiv), version 2 (see Vose et al. 2014). This dataset is of monthly time resolution and has incorporated several modern techniques to adjust data to remove biases arising from observing station inhomogeneities. It is now the standard dataset used by NOAA NCEI to assess the state of the climate in the continental United States.

Graphics illustrating daily extreme metrics of temperature and precipitation were based on NOAA NCEI's Global Historical Climatology Network-Daily (GHCN-Daily) version 3 (see Menne et al. 2012). This dataset is a comprehensive compilation of available data from climate-observing stations. Relevant to these State Climate Summaries, it includes the complete records of digital data from stations in the Cooperative Observer Network (COOP), which is the core climate network of the United States. Some stations in the COOP have observations extending back to the late 19th century. The core observations of COOP stations are daily precipitation, daily maximum temperature, daily minimum temperature, daily snowfall, and daily snow depth. The stations have been sited intentionally to provide a representative sampling of all areas of the country. The great value of this network is its longevity and spatial sampling. For this reason, it is the best observational resource to establish long-term variations and trends in the surface climate of the United States. GHCN data were also used for historical seasonal and annual temperature and precipitation analyses of Hawai‘i, Puerto Rico, and the U.S. Virgin Islands.

References:

  • Menne, M.J., I. Durre, R.S. Vose, B.E. Gleason, and T.G. Houston, 2012: An overview of the Global Historical Climatology Network-Daily database. Journal of Atmospheric and Oceanic Technology, 29, 897–910, doi:10.1175/JTECH-D-11-00103.1
  • Vose, R.S., Applequist, S., Squires, M., Durre, I., Menne, M.J., Williams, C.N., Jr., Fenimore, C., Gleason, K., & Arndt, D., 2014: Improved historical temperature and precipitation time series for U.S. climate divisions, Journal of Applied Meteorology and Climatology, 53 (5), 1232–1251. doi:10.1175/JAMC-D-13-0248.1

b. Future Climate

Projections of future climate use analyses of data from the Coupled Model Intercomparison Project Phase 5 (CMIP5; see Taylor et al. 2012). Such analyses are included in the statewide temperature time series (Figure 1: “Observed and Projected Temperature Change”) for both higher and lower emissions pathways, as well as maps (located near the end of each summary) depicting projected changes in annual or seasonal precipitation under a higher emissions pathway.

Reference:

  • Taylor, K.E., R.J. Stouffer, and G.A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bulletin of the American Meteorological Society, 93 (4), 485–498. doi:10.1175/BAMS-D-11-00094.1

CMIP5 Representative Concentration Pathways (RCPs)

The Coupled Model Intercomparison Project (CMIP) is a project of the World Climate Research Programme (WCRP) Working Group on Coupled Modeling (WGCM). This project provides a standard experimental protocol for studying Global Climate Models (GCMs). In CMIP5, 25 different modeling groups produced simulations that were used in the Intergovernmental Panel on Climate Change Fifth Assessment Report (IPCC AR5), with more than 60 representations from 28 different models.

CMIP5 includes models with higher spatial resolutions and a more developed representation of physical processes than those used for previous CMIP3 simulations. The spatial resolution of the great majority of CMIP5 simulations is in the 1-degree to 2-degree range, or about 60 to 130 miles.

CMIP5 includes the following experiments:

  1. Simulations of the 20th century using best estimates of the temporal variations in external forcing factors (such as greenhouse gas concentrations, solar output, and volcanic aerosol concentrations)
  2. Simulations of the 21st century assuming changing greenhouse gas concentrations following various scenarios

CMIP5 21st-century simulations use a set of scenarios called Representative Concentration Pathways (RCPs). These are based on radiative forcing trajectories and are named according to the radiative forcing level at the year 2100. There are four RCPs: 2.6, 4.5, 6.0, and 8.5. The numbers represent the 2100 radiative forcing increase relative to preindustrial levels in W/m2. The projected multimodel mean temperature increases at the end of the 21st century (with respect to a base period of 1901–1960) are 2.8°F, 4.2°F, 5.2°F, and 8.3°F under RCPs 2.6, 4.5, 6.0, and 8.5, respectively.

The analyses in the State Climate Summaries use simulations under RCP4.5 (a lower emissions pathway) and RCP8.5 (a higher emissions pathway). A full list of models used in these analyses can be found in the metadata accompanying each relevant figure.

For additional information on CMIP5 RCP simulations and how they compare to previous CMIP3 simulations, see Sun et al. (2015).

Reference:

  • Sun, L., K.E. Kunkel, L.E. Stevens, A. Buddenberg, J.G. Dobson, and D.R. Easterling, 2015: Regional Surface Climate Conditions in CMIP3 and CMIP5 for the United States: Differences, Similarities, and Implications for the U.S. National Climate Assessment. NOAA Technical Report NESDIS 144. National Oceanic and Atmospheric Administration, National Environmental Satellite, Data, and Information Service, 111 pp. http://dx.doi.org/10.7289/V5RB72KG


3. Analyses

Information related to the most common analyses used in the State Climate Summaries is given here. Details on additional methods used can be found in the metadata that accompany each relevant figure.

a. Choice of Reference Periods for Time Period Averages

Most of the graphics illustrate historical changes and/or future trends in climate with respect to a reference period. A common reference period was not used for all graphics. The choice of a specific reference period was determined by the purpose of the graphic and the availability of data. The great majority of graphics were based on one of two possible reference periods. The first primary reference period, 1901–1960, was used in Figure 1 (“Observed and Projected Temperature Change”) for all states in the contiguous United States, which follows the usage for several graphics in the Third National Climate Assessment (NCA3; see Melillo et al. 2014). The motivation for using this period is to illustrate climate changes occurring as a result of the recent acceleration in greenhouse gas concentrations and, specifically, how models simulate those changes. The beginning date of 1901 was chosen because earlier historical observations for the U.S. are less reliable. As shown in Meehl et al. (2003), anthropogenic forcing exhibits a slow rise during the early part of the 20th century but then accelerates after 1960, which was the reason for choosing 1960 as the ending date of the reference period. Thus, these graphics highlight changes in climate during the period of rapid increase in anthropogenic forcing and also reveal how well climate models simulate the observed changes during the more recent period of increased forcing. Due to data availability in the early part of the record, different reference periods were used for Alaska (1925–1960), Hawai‘i (1951–1980), and Puerto Rico (1951–1980).

A second primary reference period is 1971–2000, which is consistent with the World Meteorological Organization’s recommended use of 30-year periods for climate statistics. This was used for precipitation projection maps. The purpose of these graphics is to show how climate might change with respect to a period that is in people’s memory and experience. This is the same period used for precipitation projection graphics in NCA3.

The reference period for all annual and seasonal graphics (see “Statewide Temperature and Precipitation Metrics”) is simply the entire period of good observational records, which for all states in the Contiguous United States is 1895–2020. For Alaska, a reference period of 1925–2020 was used. Hawai‘i and Puerto Rico graphics use a reference period of 1950–2020.

The reference period for all threshold graphics (see “Threshold Graphics”) is also based on data availability, which varies somewhat from state to state. For the majority of states, this reference period is 1900–2020. For most of the remaining states, it is 1950–2020. A few use other periods, such as 1930–2020 and 1910–2020. Since the purpose of most of the observational graphics is to illustrate relative changes, there is no compelling rationale for an alternative reference period.

References:

  • Meehl, G.A., W.M. Washington, T.M.L. Wigley, J.M. Arblaster, A. Dai, 2003: Solar and greenhouse gas forcing and climate response in the twentieth century. Journal of Climate, 16, 426–444. doi:10.1175/1520-0442(2003)016<0426:SAGGFA>2.0.CO;2
  • Melillo, J.M., T.C. Richmond, and G.W. Yohe, Eds., 2014: Climate Change Impacts in the United States: The Third National Climate Assessment. U.S. Global Change Research Program, 841 pp. doi:10.7930/J0Z31WJ2

b. Observed and Projected Temperature Change Figures

The first figure in each summary depicts the observed and projected changes in temperature for that state and is based on Figure 17.3 in NCA3 (see Carter et al. 2014). The set of coordinated climate model simulations used in NCA4 (the Coupled Model Intercomparison Project Phase 5, or CMIP5) includes a large number of individual models. In addition, most modeling groups performed multiple simulations with their models. Each of these simulations follows a slightly different path over time because of natural variability in the climate system, which varies between models and within the multiple simulations of a single model (see “Datasets”). The path of annual temperature in a specific simulation represents this random natural variability, as well as the known increasing greenhouse gas forcing. Because of this random natural variability, there is no expectation that a model simulation will exactly match the observed annual temperature path. However, with a large number of simulations, we can examine statistically whether the model simulations bound the actual observations. The purpose of Figure 1 is to illustrate this by comparing the observed annual temperature with the entire distribution of climate model simulations for the historical period. In addition, this analysis also shows the uncertainty of the future evolution of climate out to 2100.

In Figure 1, the statewide-average observed annual temperature is displayed at annual time resolution. As noted in “Choice of Reference Periods for Time Period Averages” each annual value is expressed as the difference between annual temperature and the 1901–1960 annual average temperature. In order to provide an apples-to-apples comparison, the model simulations are summarized at the annual resolution. The details of the model simulation analysis are presented with the figure metadata. A couple of points are particularly noteworthy. First, the model-simulated statewide average temperature values are also expressed as the difference between the annual temperature and the model’s 1901–1960 average. Second, the shading in Figure 1 indicates the spread of the model-simulated annual values, specifically the 5th to 95th percentile range. For each year, the set of individual model values are ranked and then the 5th and 95th percentile values are determined.

Reference:

  • Carter, L.M., J.W. Jones, L. Berry, V. Burkett, J.F. Murley, J. Obeysekera, P.J. Schramm, and D. Wear, 2014: Southeast and the Caribbean. In Climate Change Impacts in the United States: The Third National Climate Assessment. Melillo, J.M., T.C. Richmond, and G.W. Yohe, Eds., U.S. Global Change Research Program, Washington, DC, 396–417. doi:10.7930/J0NP22CB

c. Statewide Temperature and Precipitation Metrics

Use of 5-year Averages

The majority of figures in the summaries are time series of various temperature and precipitation metrics. These charts are based on annual values, which are displayed as a line graph. In addition, 5-year averages are plotted as bars. The choice of whether to average, and how long to average, is a compromise between the clear and uncluttered presentation of long-term variations and any trends versus the inclusion of all available details. While the graphical presentation of the underlying annual data has the advantage of illustrating year-to-year variability, it can obscure any longer-term behavior. Since the main focus of the summaries is on longer-term behavior, we opted to apply block averages to provide a clear presentation of this aspect of the time series and to present those averages as bars, making them the visual center of the graphic. At the same time, we included the annual values as an overlying line graph to provide the full details of the data. The periods of analysis (e.g., 1895–2020) are not integer multiples of 5. To include all data in the bars, the last period of averaging is 6 years: 2015–2020. Because of how seasons are defined (see “Definition of Seasons”), the first bar for all winter graphics is an average of only 4 years (1896–1899).

Calculations of Temperature and Precipitation at Various Timescales

The daily maximum temperature (Tmax) is the highest temperature occurring in the 24 hours prior to the daily observation time. Likewise, the daily minimum temperature (Tmin) is the lowest temperature occurring in the 24 hours prior to the daily observation time. The daily average temperature (Tavg) is calculated from Tmax and Tmin simply as Tavg = (Tmax + Tmin)/2.

Seasonal and annual temperature values are calculated from Tavg, Tmax, and Tmin. The seasonal and annual values are derived from calculated monthly averages. Monthly average temperature Tmonavg is the average of Tavg for all days in the month. Monthly average maximum temperature Tmonmax is the average of Tmax for all days in the month. Monthly average minimum temperature Tmonmin is the average of Tmin for all days in the month. Annual average temperature is the average of the 12 values of Tmonavg. Annual average maximum temperature is the average of the 12 values of Tmonmax. Annual average minimum temperature is the average of the 12 values of Tmonmin. Seasonal average temperature is calculated as the average of the 3 monthly values of Tmonavg in that season (December, January, and February for winter; March, April, and May for spring; June, July, and August for summer; September, October, and November for fall). Seasonal average maximum temperature is calculated as the average of the 3 monthly values of Tmonmax in that season. Seasonal average minimum temperature is calculated as the average of the 3 monthly values of Tmonmin in that season.

Daily precipitation is the total amount of precipitation occurring in the 24 hours prior to the daily observation time. Seasonal and annual precipitation values are calculated as the sum of monthly precipitation amounts.

Definition of Seasons

The meteorological/climatological definitions of the seasons are used in the summaries. The seasons are defined around whole calendar months as follows:

  • Winter: December 1–February 28/29
  • Spring: March 1–May 31
  • Summer: June 1–August 31
  • Fall: September 1–November 30

NOAA NCEI’s Climate Divisional Dataset (nClimDiv), version 2 (see Vose et al. 2014) is used in the summaries for most graphics that display average temperature and precipitation. The period of record of this dataset is 1895–present. Since the dataset does not include December 1894, the first full winter that is available is December 1, 1895, to February 29, 1896, denoted as winter 1896. Thus, winter season graphics begin with the year 1896 while all other seasons begin with 1895.

Reference:

  • Vose, R.S., Applequist, S., Squires, M., Durre, I., Menne, M.J., Williams, C.N., Jr., Fenimore, C., Gleason, K., & Arndt, D., 2014: Improved historical temperature and precipitation time series for U.S. climate divisions, Journal of Applied Meteorology and Climatology, 53 (5), 1232–1251. doi:10.1175/JAMC-D-13-0248.1

Threshold Graphics

Each state summary includes a number of graphics that display the annual number of days that daily temperature or precipitation is above or below a selected threshold. These were developed from individual stations in NOAA NCEI’s Global Historical Climatology Network-Daily (GHCN-Daily). The climate-observing stations were selected based on data availability. Specifically, only stations with less than 10% missing data for the period of analysis were used. This approach was adopted to ensure that the mix of stations was relatively static throughout the period of analysis; otherwise, it would have been possible to introduce artificial trends because of stations dropping in and out.

The detailed procedures used to produce these graphics are provided in the metadata that are associated with each figure. A few general points are provided here with regard to the motivation for certain aspects of the methodology. The values that are plotted should be viewed as an index, with the absolute numbers representing a typical station for that state. The method used to produce the graphed index recognizes that the climatology of the specific variable can vary considerably across a state, particularly for those states with substantial topographic variations. For example, for the 1991–2020 period, the number of days with daily maximum temperature of 100°F or higher in Arizona is 111 at Phoenix (with a range of 88–145), 15 at Tombstone (with a range of 2–42), and only 1 at Prescott (with a range of 0–7). Without considering such differences, the year-to-year variations at Phoenix could negate potentially opposite behavior at Tombstone and Prescott. Another potential issue is an uneven distribution of stations. In large states in particular (e.g., California and Texas), the temporal trends and variations could be different across regions. If one region of the state had fewer stations than other regions, the results could be skewed. To address these two specific issues, the following general approaches were adopted:

  1. In order to incorporate all stations equally into the computed index, the station time series of annual values were converted from numbers of days into standardized anomalies. These were calculated by subtracting the station’s time series mean and dividing by the station’s time series standard deviation for each annual count for that station. This resulted in a time series with a mean of zero. All of the averaging to produce a statewide number was done using the standardized anomaly time series. Once the statewide averaging was completed (see step 2 below for how that was done), which resulted in a state-average standardized anomaly time series, the values were converted back to real numbers. This was done by first converting the standardized anomalies to numbers of days by multiplying the standardized anomaly by the mean of the station standard deviations. Then, this anomaly value was added to the station average of mean values of numbers of days to obtain a value of number of days for each year.

  2. In order to minimize unequal weighting of different regions of a state, we did not simply average all of the stations together. Instead, a state was divided into 1-degree grid boxes. Then, grid box time series were produced by averaging all stations within that grid box. Finally, a state-average time series was created by averaging the grid box time series.

Reference:

  • Menne, M.J., I. Durre, R.S. Vose, B.E. Gleason, and T.G. Houston, 2012: An overview of the Global Historical Climatology Network-Daily database. Journal of Atmospheric and Oceanic Technology, 29, 897–910, doi:10.1175/JTECH-D-11-00103.1

d. Precipitation Projections Maps

Each state summary includes a map depicting projected changes in annual or seasonal precipitation between the middle of the 21st century (2041–2070) and the climate model reference period (1971–2000) under a higher emissions pathway using CMIP5 data. These maps include an indication of model agreement, determined using the following technique:

The statistical significance regarding the change in precipitation was determined using a 2-sample t-test, assuming unequal variances for those two samples. For each period (present and future climate), the mean and standard deviation were calculated using the 30 annual values. These were then used to calculate t. In order to assess the agreement between models, the following three categories were determined for each grid point:

  • Category 1: If less than 50% of the models indicate a statistically significant change, then the multimodel mean is shown in color. This means that model results are in general agreement that simulated changes are within historical variations.
  • Category 2: If more than 50% of the models indicate a statistically significant change but less than 67% of the significant models agree on the sign of the change, then the grid points are whited out, indicating that the models are in disagreement about the direction of change.
  • Category 3: If more than 50% of the models indicate a statistically significant change and more than 67% of the significant models agree on the sign of the change, then the multimodel mean is shown in color with hatching. Model results are in agreement that simulated changes are statistically significant and in a particular direction.

More information, as well as U.S. precipitation maps for additional emissions pathways and time periods, can be found in the following report:

  • Sun, L., K.E. Kunkel, L.E. Stevens, A. Buddenberg, J.G. Dobson, and D.R. Easterling, 2015: Regional Surface Climate Conditions in CMIP3 and CMIP5 for the United States: Differences, Similarities, and Implications for the U.S. National Climate Assessment. NOAA Technical Report NESDIS 144. National Oceanic and Atmospheric Administration, National Environmental Satellite, Data, and Information Service, 111 pp. http://dx.doi.org/10.7289/V5RB72KG

e. Tidal Flooding Figures

Figures depicting the number of observed and projected tidal flood days are included in the summaries for most coastal states. Observed data for 1920–2020 were derived from counting the number of days per year when water levels exceed the derived NOAA threshold for minor (high tide) flooding (see Sweet et al. 2018). A definition of “year” used for these analyses is May–April so as not to divide the winter season (this is important to account for variability in the El Niño–Southern Oscillation). Only years with 80% or more data completion were used. Tidal flood projections for 2021–2100 are based on federal interagency global sea level rise scenarios for the United States, which are projected onto a 1-degree grid for the entire U.S. shoreline and include additional regional sea level changes that result from changes in land elevation, earth’s gravitational field and rotation, and ocean circulation to project changes in high tide flood frequencies. Projections are included under the Intermediate-Low and Intermediate scenarios (see Sweet et al., 2017). These scenarios correspond to a global mean sea level rise of 0.5 m and 1.0 m, respectively, and are consistent with those used in NOAA’s annual High Tide Flooding Report.

References: