ArticlePublisher preview available

Improvements in the uncertainty model in the Goddard Institute for Space Studies Surface Temperature (GISTEMP) analysis

  • NASA Goddard Institute for Space Studies,
To read the full-text of this research, you can request a copy directly from the authors.

Abstract and Figures

We outline a new and improved uncertainty analysis for the Goddard Institute for Space Studies Surface Temperature product version 4 (GISTEMP v4). Historical spatial variations in surface temperature anomalies are derived from historical weather station data and ocean data from ships, buoys, and other sensors. Uncertainties arise from measurement uncertainty, changes in spatial coverage of the station record, and systematic biases due to technology shifts and land cover changes. Previously published uncertainty estimates for GISTEMP included only the effect of incomplete station coverage. Here, we update this term using currently available spatial distributions of source data, state‐of‐the‐art reanalyses, and incorporate independently derived estimates for ocean data processing, station homogenization, and other structural biases. The resulting 95% uncertainties are near 0.05 °C in the global annual mean for the last 50 years and increase going back further in time reaching 0.15 °C in 1880. In addition, we quantify the benefits and inherent uncertainty due to the GISTEMP interpolation and averaging method. We use the total uncertainties to estimate the probability for each record year in the GISTEMP to actually be the true record year (to that date) and conclude with 87% likelihood that 2016 was indeed the hottest year of the instrumental period (so far).
This content is subject to copyright. Terms and conditions apply.
Improvements in the GISTEMP Uncertainty Model
Nathan J. L. Lenssen1,2 , Gavin A. Schmidt1, James E. Hansen3, Matthew J. Menne4,
Avraham Persin1,5, Reto Ruedy1,5, and Daniel Zyss1
1NASA Goddard Institute for Space Studies, New York, NY, USA, 2Department of Earth and Environmental Sciences,
Columbia University, New York, NY, USA, 3Climate Science, Awareness and Solutions, Columbia University Earth
Institute, New York, NY, USA, 4NOAA National Centers for Environmental Information, Asheville, NC, USA, 5SciSpace
LLC, New York, NY, USA
Abstract We outline a new and improved uncertainty analysis for the Goddard Institute for Space
Studies Surface Temperature product version 4 (GISTEMP v4). Historical spatial variations in surface
temperature anomalies are derived from historical weather station data and ocean data from ships, buoys,
and other sensors. Uncertainties arise from measurement uncertainty, changes in spatial coverage of the
station record, and systematic biases due to technology shifts and land cover changes. Previously published
uncertainty estimates for GISTEMP included only the effect of incomplete station coverage. Here, we
update this term using currently available spatial distributions of source data, state-of-the-art reanalyses,
and incorporate independently derived estimates for ocean data processing, station homogenization, and
other structural biases. The resulting 95% uncertainties are near 0.05 C in the global annual mean for the
last 50 years and increase going back further in time reaching 0.15 C in 1880. In addition, we quantify the
benefits and inherent uncertainty due to the GISTEMP interpolation and averaging method. We use the
total uncertainties to estimate the probability for each record year in the GISTEMP to actually be the true
record year (to that date) and conclude with 8 % likelihood that 2016 was indeed the hottest year of the
instrumental period (so far).
1. Introduction
Attempts to seriously estimate the changes in temperature at the hemispheric and global scale date back at
least to Callendar (1938) who used 147 land-based weather stations to track near-global trends from 1880
to 1935 (Hawkins & Jones, 2013). Subsequent efforts used substantially more data (180 stations in Mitchell,
1961; 400 stations in Callendar, 1961; “several hundred” in Hansen et al., 1981; etc.), and with a greater
global reach. While efforts were made to estimate the uncertainty associated with these products, they were
more suggestive than comprehensive.
As the data sets have grown in recent years (through digitization and synthesis of previously separate data
streams; Freeman et al., 2016; Rennie et al., 2014; Thorne et al., 2018), and efforts have been made to improve
data homogenization, bias corrections, and interpolation schemes, the sophistication of the uncertainty
models has also grown. Notably, with the introduction of the Hadley Centre sea surface temperature (SST)
analysis HadSST3 (Kennedy et al., 2011a, 2011b), Berkeley Earth (Rohde et al., 2013a), and the joint Hadley
Centre and University of East Anglia's Climatic Research Unit Hadley Centre/Climatic Research Unit 4
(HadCRUT4; Morice et al., 2012), Monte Carlo methodologies have been applied to generate observational
ensembles that quantify uncertainties more comprehensively than was previously possible.
Goddard Institute for Space Studies Surface Temperature (GISTEMP) is a widely used data product that
tracks global climate change over the instrumental era. However, the existing uncertainty analysis cur-
rently contains only rough estimates of uncertainty on the land surface air temperature (LSAT) mean and
no estimates of the SST or total (land and sea surface combined) global mean. This paper describes a new
end-to-end assessment of all the known uncertainties associated with the current GISTEMP analysis (nom-
inally based in the methodology described in Hansen et al., 2010, but with changes to data sources as
documented on the GISTEMP website and outlined below), denoted as version 4. We use independently
derived uncertainty models for the land station homogenization (Menne et al., 2010, 2018) and ocean tem-
perature products (Huang et al., 2015, 2017), combined with our own assessment of spatial interpolation
and coverage uncertainties, as well as parametric uncertainty in the GISTEMP methodology itself.
Key Points:
• A total uncertainty analysis for
GISTEMP is presented for the first
• Uncertainty in global mean surface
temperature is roughly 0.05
degrees Celsius in recent decades
increasing to 0.15 degrees Celsius in
the nineteenth century
• Annual mean uncertainties are small
relative to the long-term trend
Correspondence to:
N. J. L. Lenssen,
Lenssen, N. J. L., Schmidt, G. A.,
Hansen, J. E., Menne, M. J., Persin, A.,
Ruedy, R., & Zyss, D. (2019).
ImpImprovements in the GISTEMP
uncertainty model. Journal of
Geophysical Research: Atmospheres,
124, 6307–6326.
Received 23 AUG 2018
Accepted 12 MAY 2019
Accepted article online 23 MAY 2019
Published online 24 JUN 2019
Corrected 27 AUG 2019 and
28 FEB 2020
This article was corrected on
2019 and 28 FEB 2020. See the end
of the full text for details.
©2019. American Geophysical Union.
All Rights Reserved.
27 AUG
Full-text available
Satellite-based normalized difference vegetation index (NDVI) time series data are useful for monitoring the changes in vegetation ecosystems in the context of global climate change. However, most of the current NDVI products cannot effectively reconcile high spatial resolution and continuous observations in time. Here, to produce a global-scale, long-term, and high-resolution NDVI database, we developed a simple and new data downscaling approach. The downscaling algorithm considers the pixel-wise ratios of the coefficient of variation (CV) between the coarse-and fine-resolution NDVI data and relative changes in the NDVI against a baseline period. The algorithm successfully created a worldwide monthly NDVI database with 250 m resolution from 1982 to 2018 by translating the fine spatial information from MODIS (Moderate-resolution Imaging Spec-troradiometer) data and the long-term temporal information from AVHRR (Advanced Very High Resolution Radiometer) data. We employed the evaluation indices of root mean square error (RMSE), mean absolute error (MAE), and Pearson's correlation coefficient (Pearson's R) to assess the accuracy of the downscaled data against the MODIS NDVI. Both the RMSE and MAE values at the regional and global scales are typically between 0 and 0.2, whereas the Pearson's R values are mostly above 0.7, which implies that the downscaled NDVI product is similar to the MODIS NDVI product. We then used the downscaled data to monitor the NDVI changes in different plant types and places with significant vegetation heterogeneity, as well as to investigate global vegetation trends over the last four decades. The Google Earth Engine platform was used for all the data downscaling processes, and here we provide a code for users to easily acquire data corresponding to any part of the world. The downscaled global-scale NDVI time series has high potential for the monitoring of the long-term temporal and spatial dynamics of terrestrial ecosystems under changing environments.
The mechanisms underlying decadal variability in Arctic sea ice remain actively debated. Here, we show that variability in boreal biomass burning (BB) emissions strongly influences simulated Arctic sea ice on multidecadal time scales. In particular, we find that a strong acceleration in sea ice decline in the early 21st century in the Community Earth System Model version 2 (CESM2) is related to increased variability in prescribed BB emissions in the sixth phase of the Coupled Model Intercomparison Project (CMIP6) through summertime aerosol-cloud interactions. Furthermore, we find that more than half of the reported improvement in sea ice sensitivity to CO2 emissions and global warming from CMIP5 to CMIP6 can be attributed to the increased BB variability, at least in the CESM. These results highlight a new kind of uncertainty that needs to be considered when incorporating new observational data into model forcing while also raising questions about the role of BB emissions on the observed Arctic sea ice loss.
Full-text available
This publication analyzes changes in global warming rates (GWR) expressed in Centigrade per year (°C/y) and introduces a parameter “global warming acceleration” (GWA) expressed in Centigrade per year per year (°C/y2). GWA may be applied to forecast the GWR for the next decade. If the current decrease of GWA will continue for the next 11 years, Global Warming Rate will increase from the current +0.017 °C/y to +0.019 °C/y, for land+ocean.
Full-text available
This publication analysis a correlation between the cumulative CO2 emissions and the global surface temperature change. The change in global surface temperature was calculated as the difference in the two most recent 31 years neighboring periods, and the change in the cumulative CO2 emissions is the difference between the actual values in the centers of the periods. This approach resulted in 0.000556°C/GtCO2 for the year 2000 and 0.000745°C/GtCO2 for the year 2021.
Rainfall forecasting is essential to manage water resources and make timely decisions to mitigate adverse effects related to unexpected events. Considering that rainfall drivers can change throughout the year, one approach to implementing forecasting models is to generate a model for each period in which the mechanisms are nearly constant, e.g., each season. The chosen predictors can be more robust, and the resulting models perform better. However, it has not been assessed whether the approach mentioned above offers better performance in forecasting models from a practical perspective in the tropical Andean region. This study evaluated quarterly, semiannual and annual models for forecasting monthly rainfall anomalies in an Andean basin to show if models implemented for fewer months outperform accuracy; all the models forecast rainfall on a monthly scale. Lagged rainfall and climate indices were used as predictors. Support vector regression (SVR) was used to select the most relevant predictors and train the models. The results showed a better performance of the annual models mainly due to the greater amount of data that SVR can take advantage of in training. If the training of the annual models had less data, the quarterly models would be the best. In conclusion, the annual models show greater accuracy in the rainfall forecast.
The demand for meteorological gridded datasets has increased within the last few years to inform studies such those in climate, weather, and agriculture. These studies require those data to be readily usable in standard formats with continuous spatial and temporal coverage. Since 2016, Brazil has a daily gridded meteorological data set with spatial resolution of 0.25°× 0.25° from 01/01/1980 to 31/12/2013 which was well received by the community. The main objective of this work is to improve the Brazilian meteorological data set. We do this by increasing the resolution of the minimum and maximum temperature (Tmax and Tmin) gridded interpolations from 0.25°× 0.25° to 0.1°× 0.1° by incorporating data on topographic relief, and increasing the time period covered (01/01/1961‐07/31/2020). Besides Tmax and Tmin, we also gridded precipitation (pr), solar radiation (Rs), wind speed (u2), and relative humidity (RH) using observed data from 11, 473 rain gauges and 1, 252 weather stations. By means of ranked cross‐validation statistics, we selected the best interpolation among inverse distance weighting (IDW) and angular distance weighting (ADW) methods. We determined that interpolations for Tmax and Tmin are improved by using the elevation of a query point, that accounts for topographic relief, and a temperature lapse rate. Even though this new version has ≈ 25 years more data relative to the previous one, statistics from cross‐validation were similar. To allow researchers to assess the performance of the interpolation relative to station data in the area, we provide two types of gridded controls. This article is protected by copyright. All rights reserved.
Full-text available
Plain Language Summary Long‐term observations reveal that Earth's average temperature is rising due to human‐caused climate change. Along with this warming trend are also variations from year‐to‐year and even over multiple decades. This temperature variability is often tied to regional patterns of heat in the deep ocean, which can then modulate weather and climate extremes over land. In an attempt to better predict temperature variability on decadal timescales, we use a machine learning method called artificial neural networks and data from a climate model experiment, which was designed to compare climate change and variability. Here, our artificial neural network (ANN) uses maps of ocean heat to predict the onset of temporary slowdowns in the rate of global warming in both the climate model and in real‐world observations. We then use a visualization technique to find which areas of ocean heat that the ANN is using to make its correct predictions, which are found to be mainly across the Pacific Ocean. In agreement with recent research, our study finds that new data science methods, like machine learning, can be useful tools for predicting variations in global climate.
Technical Report
Full-text available
This publication includes details regarding the online converter of global warming baselines "Global Warming Online Baselines Converter, °C" available on the Internet site This converter displays the conversion factor from any period to any period in the range 1850-2021 for land+ocean, 1880-2021 for ocean-only, or 1750-2021 for land-only.
This study involves examination of glaciological mass-balance time series, glacier and climatic descriptors, the application of machine learning methods for glaciological clustering, and computation of mass-balance time series based upon the clustering and statistical analyses relative to gridded air temperature datasets. Our analysis revealed an increasingly coherent mass-balance trend but a latitudinal bias of monitoring programs. The glacier classification scheme delivered three clusters, suggesting these correspond to climate-based first-order regimes, as glacier morphometric characteristics weighed little in our multivariate analysis. We combined all available surface mass-balance data from in situ monitoring programs to study temperature sensitivity for each cluster. These aggregated mass-balance time series delivered spatially different statistical relationships to temperature. Results also showed that surface mass balance tends to have a temporal self-correlation of ~20 years. Using this temporal window to analyze sensitivity since ~ 1950, we found that in all cases temperature sensitivity, while generally negative, tended to fluctuate through time, with the largest absolute magnitudes occurring in the 1980s and becoming less negative in recent years, revealing that glacier sensitivity is non-stationary. These findings point to a scenario of a coherent signal of change no matter the glacier regime. This work provides new insights into glacier–climate relationships that can guide observational and analytical strategies.
Full-text available
This paper presents Atmospheric Infra-Red Sounder (AIRS) surface skin temperature anomalies for the period 2003 through 2017, and compares them to station-based analyses of surface air temperature anomalies (principally the Goddard Institute for Space Studies Surface Temperature Analysis (GISTEMP)). The AIRS instrument flies on EOS Aqua, which was launched in 2002 and became stable in September 2002. AIRS surface temperatures are completely satellite-based and are totally independent of any surface-based measurements. We show in this paper that satellite-based surface temperatures can serve as an important validation of surface-based estimates and help to improve surface-based data sets in a way that can be extended back many decades to further scientific research. AIRS surface temperatures have better spatial coverage than those of GISTEMP, though at the global annual scale the two data sets are highly coherent. As in the surface-based analyses, 2016 was the warmest year yet.
Full-text available
There is overwhelming evidence that the climate system has warmed since the instigation of instrumental meteorological observations. The Fifth Assessment Report of the Intergovernmental Panel on Climate Change concluded that the evidence for warming was unequivocal. However, owing to imperfect measurements and ubiquitous changes in measurement networks and techniques, there remain uncertainties in many of the details of these historical changes. These uncertainties do not call into question the trend or overall magnitude of the changes in the global climate system. Rather, they act to make the picture less clear than it could be, particularly at the local scale where many decisions regarding adaptation choices will be required, both now and in the future. A set of high-quality long-term fiducial reference measurements of essential climate variables will enable future generations to make rigorous assessments of future climate change and variability, providing society with the best possible information to support future decisions. Here we propose that by implementing and maintaining a suitably stable and metrologically well-characterized global land surface climate fiducial reference measurements network, the present-day scientific community can bequeath to future generations a better set of observations. This will aid future adaptation decisions and help us to monitor and quantify the effectiveness of internationally agreed mitigation steps. This article provides the background, rationale, metrological principles, and practical considerations regarding what would be involved in such a network, and outlines the benefits which may accrue. The challenge, of course, is how to convert such a vision to a long-term sustainable capability providing the necessary well-characterized measurement series to the benefit of global science and future generations.
Full-text available
The United Nations Framework Convention on Climate Change (UNFCCC) process agreed in Paris to limit global surface temperature rise to ‘well below 2°C above pre-industrial levels’. But what period is ‘pre-industrial’? Some-what remarkably, this is not defined within the UNFCCC’s many agreements and protocols. Nor is it defined in the IPCC’s Fifth Assessment Report (AR5) in the evaluation of when particular temperature levels might be reached because no robust definition of the period exists. Here we discuss the important factors to consider when defining a pre-industrial period, based on estimates of historical radiative forcings and the availability of climate observations. There is no perfect period, but we suggest that 1720-1800 is the most suitable choice when discussing global temperature limits. We then estimate the change in global average temperature since pre-industrial using a range of approaches based on observations, radiative forcings, global climate model simulations and proxy evidence. Our assessment is that this pre-industrial period was likely 0.55–0.80°C cooler than 1986-2005 and that 2015 was likely the first year in which global average temperature was more than 1°C above pre-industrial levels. We provide some recommendations for how this assessment might be improved in future and suggest that reframing temperature limits with a modern baseline would be inherently less uncertain and more policy-relevant.
We describe a fourth version of the Global Historical Climatology Network (GHCN)-monthly (GHCNm) temperature dataset. Version 4 (v4) fulfills the goal of aligning GHCNmtemperature values with the GHCNdaily dataset and makes use of data from previous versions of GHCNm as well as data collated under the auspices of the International Surface Temperature Initiative. GHCNm v4 has many thousands of additional stations compared to version 3 (v3) both historically and with short time-delay updates. The greater number of stations as well as the use of records with incomplete data during the base period provides for greater global coverage throughout the record compared to earlier versions. Like v3, the monthly averages are screened for random errors and homogenized to address systematic errors. New to v4, uncertainties are calculated for each station series, and regional uncertainties scale directly from the station uncertainties. Correlated errors in the station series are quantified by running the homogenization algorithm as an ensemble. Additional uncertainties associated with incomplete homogenization and use of anomalies are then incorporated into the station ensemble. Further uncertainties are quantified at the regional level, the most important of which is for incomplete spatial coverage. Overall, homogenization has a smaller impact on the v4 global trend compared to v3, though adjustments lead to much greater consistency than between the unadjusted versions. The adjusted v3 global mean therefore falls within the range of uncertainty for v4 adjusted data. Likewise, annual anomaly uncertainties for the other major independent land surface air temperature datasets overlap with GHCNm v4 uncertainties.
Several groups routinely produce gridded land surface air temperature (LSAT) data sets using station measurements to assess the status and impact of climate change. The Intergovernmental Panel on Climate Change Fifth Assessment Report suggests that estimated global and hemispheric mean LSAT trends of different data sets are consistent. However, less attention has been paid to the intercomparison at local/regional scales, which is important for local/regional studies. In this study we comprehensively compare four data sets at different spatial and temporal scales, including Berkley Earth Surface Temperature land surface air temperature data set (BEST-LAND), Climate Research Unit Temperature Data Set version 4 (CRU-TEM4v), National Aeronautics and Space Administration Goddard Institute for Space Studies data (NASA-GISS), and data provided by National Oceanic and Atmospheric Administration National Center for Environmental Information (NOAA-NCEI). The mean LSAT anomalies are remarkably different because of the data coverage differences, with the magnitude nearly 0.4°C for the global and Northern Hemisphere and 0.6°C for the Southern Hemisphere. This study additionally finds that on the regional scale, northern high latitudes, southern middle-to-high latitudes, and the equator show the largest differences nearly 0.8°C. These differences cause notable differences for the trend calculation at regional scales. At the local scale, four data sets show significant variations over South America, Africa, Maritime Continent, central Australia, and Antarctica, which leads to remarkable differences in the local trend analysis. For some areas, different data sets produce conflicting results of whether warming exists. Our analysis shows that the differences across scales are associated with the availability of stations and the use of infilling techniques. Our results suggest that conventional LSAT data sets using only station observations have large uncertainties across scales, especially over station-sparse areas. In developing future LSAT data sets, the data uncertainty caused by limited and unevenly distributed station observations must be reduced.
The monthly global 2° × 2° Extended Reconstructed Sea Surface Temperature (ERSST) has been revised and updated from version 4 to version 5. This update incorporates a new release of ICOADS release 3.0 (R3.0), a decade of near-surface data from Argo floats, and a new estimate of centennial sea ice from HadISST2. A number of choices in aspects of quality control, bias adjustment, and interpolation have been substantively revised. The resulting ERSST estimates have more realistic spatiotemporal variations, better representation of high-latitude SSTs, and ship SST biases are now calculated relative to more accurate buoy measurements, while the global long-term trend remains about the same. Progressive experiments have been undertaken to highlight the effects of each change in data source and analysis technique upon the final product. The reconstructed SST is systematically decreased by 0.077°C, as the reference data source is switched from ship SST in ERSSTv4 to modern buoy SST in ERSSTv5. Furthermore, high-latitude SSTs are decreased by 0.1°-0.2°C by using sea ice concentration from HadISST2 over HadISST1. Changes arising from remaining innovations are mostly important at small space and time scales, primarily having an impact where and when input observations are sparse. Cross validations and verifications with independent modern observations show that the updates incorporated in ERSSTv5 have improved the representation of spatial variability over the global oceans, the magnitude of El Niño and La Niña events, and the decadal nature of SST changes over 1930s-40s when observation instruments changed rapidly. Both long- (1900-2015) and short-term (2000-15) SST trends in ERSSTv5 remain significant as in ERSSTv4.
The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2), is the latest atmospheric reanalysis of the modern satellite era produced by NASA's Global Modeling and Assimilation Office (GMAO). MERRA-2 assimilates observation types not available to its predecessor, MERRA, and includes updates to the Goddard Earth Observing System (GEOS) model and analysis scheme so as to provide a viable ongoing climate analysis beyond MERRA's terminus. While addressing known limitations of MERRA, MERRA-2 is also intended to be a development milestone for a future integrated Earth system analysis (IESA) currently under development at GMAO. This paper provides an overview of the MERRA-2 system and various performance metrics. Among the advances in MERRA-2 relevant to IESA are the assimilation of aerosol observations, several improvements to the representation of the stratosphere including ozone, and improved representations of cryospheric processes. Other improvements in the quality of MERRA-2 compared with MERRA include the reduction of some spurious trends and jumps related to changes in the observing system and reduced biases and imbalances in aspects of the water cycle. Remaining deficiencies are also identified. Production of MERRA-2 began in June 2014 in four processing streams and converged to a single near-real-time stream in mid-2015. MERRA-2 products are accessible online through the NASA Goddard Earth Sciences Data Information Services Center (GES DISC).
The ERA-Interim and JRA-55 reanalyses of synoptic data and several conventional analyses of monthly climatological data provide similar estimates of global-mean surface warming since 1979. They broadly agree on the character of interannual variability and the extremity of the 2015/16 warm spell to which a strong El Niño and low Arctic sea-ice cover contribute. Global and regional averages nevertheless differ on various timescales due to differences in data coverage and sea-surface temperature analyses; averages from those conventional datasets that infill where they lack direct observations agree better with the averages from the reanalyses. The latest warm event is less extreme when viewed in terms of atmospheric energy, which gives more weight to variability in the tropics, where the thermal signal has greater vertical penetration and latent energy is a larger factor.
We highlight improvements to the International Comprehensive Ocean-Atmosphere Data Set (ICOADS) in the latest Release 3.0 (R3.0; covering 1662–2014). ICOADS is the most widely used freely available collection of surface marine observations, providing data for the construction of gridded analyses of sea surface temperature, estimates of air–sea interaction and other meteorological variables. ICOADS observations are assimilated into all major atmospheric, oceanic and coupled reanalyses, further widening its impact. R3.0 therefore includes changes designed to enable effective exchange of information describing data quality between ICOADS, reanalysis centres, data set developers, scientists and the public. These user-driven innovations include the assignment of a unique identifier (UID) to each marine report – to enable tracing of observations, linking with reports and improved data sharing. Other revisions and extensions of the ICOADS' International Maritime Meteorological Archive common data format incorporate new near-surface oceanographic data elements and cloud parameters. Many new input data sources have been assembled, and updates and improvements to existing data sources, or removal of erroneous data, made. Coupled with enhanced ‘preliminary’ monthly data and product extensions past 2014, R3.0 provides improved support of climate assessment and monitoring, reanalyses and near-real-time applications.