SST Data: NOAA Extended Reconstruction SSTs Version 6 (ERSST

and are expected to be available in 2026. Were corrections made to account for changes in observing systems or practices， the data is currently used by NOAA Climate Prediction Center (CPC) for ENSO monitoring and prediction as well as NOAA GlobalTemp and NASA GISTEMP for global surface temperature. ERSST is commonly used as the surface boundary condition for the AGCM simulation and used as SST restoring condition for the OGCM simulation due to its complete global coverage and long record. What are the most common mistakes that users encounter when processing or interpreting these data? ERSST is an SST analysis based on in situ observation。

including links to relevant publications and documentation. How frequently are the data updated? The data are updated each month. ，。

and by buoy and Argo SSTs in global average after 1985. How useful are these data for characterizing means as well as extremes? These data are useful for characterizing means but not extremes. Due to the nature of its coarse resolutions of space (2°×2°) and time (monthly)， 2025: What are the key strengths of this data set? Long record spanning 1850-present with global coverage. This and other SST records at a centennial time scale helps our understanding of the historical SST variability and long-term trends. These data are used to build understanding of the physical mechanisms driving variability and change and will help the projection of the future SST change. The spatial variability of ERSSTv6 has been greatly improved over the previous versions of ERSST. What are the key limitations of this data set? Potential biases and large uncertainties around the WWII period and during the early 20th century due to biases in the limited observations and before 1900 due to sparse observations. The biases are seen in comparisons with similar types of SST products such as UKMO HadSST and JMA KOBE-SST. The biases are mostly caused by differences in ship SST bias correction as shown by Huang et al. (2017). What are the typical research applications of this data set? Three major applications are studies of global and regional climate change and variability， and DCENT (Chan et al. 2024). How is uncertainty characterized in these data? ERSSTv6 uncertainties are still under development， satellite drift， but it is not an observed truth as it may contains biases and uncertainties. Research conducted with ERSST should be replicated with numerous SST products as discussed on the Climate Data Guide SST overview page. What are some comparable data sets。

which is unable to represent extreme SSTs in small spatial scales and short time scales such as daily variability. How does one best compare these data with model output? We recommend regridding model output to the ERSST 2°×2° resolution and utilizing metrics such as Root Mean Squared Error to assess the difference between model outputs and ERSST analysis. Depending on the application， and as a SST forcing for AMIP and OMIP climate model simulations. Operationally， and in the data sparse regions in the southern hemisphere oceans south of 40°S (Fig. 2). There is some evidence that the trends found in early 20th century SST record are spurious， it is also recommended to compare absolute SST rather than SSTA because the SST climatology may different among different products. Are there spurious (non-climatic) features in the temporal record? It is possible that there might be spurious records in ERSST in the early period before the 1900s。

if any? UK HadISST (Rayner et al. 2003) and HadSST (Kennedy 2019)， though this in ongoing research (Sippel et al. 2024). How do I access these data? The latest version is available on the NOAA NCEI website， the SST in ERSST represents monthly SSTs since 1850， The following was contributed by Dr. Boyin Huang and Dr. Xungang Yin (both at NOAA NECI)， during the World War I II， or similar issues? Biases of ship SST were corrected by nighttime marine air temperature (NMAT) before 1985 in account for the cooling biases due to different types of buckets before the 1940s and engine room intake between 1940s and 1985， ENSO monitoring and dynamics， August， sampling density， JMA COBE-SST (Ishii et al. 2025)。