Disclaimer:
These data are considered PRELIMINARY. While qualitative control methods were utilized to scrutinize the temperature data, erroneous, late arriving and/or missed reports are possible. As such, the data have not undergone final quality control by the
National Centers for Environmental Information (NCEI). Therefore, these data are subject to revision. Final and certified climate data can be accessed at
NCEI.
What is this product?
The Weather Prediction Center (WPC) National High and Low Temperature product as seen on this webpage (or disseminated XML file as either, XTUES or WMO header RXUS30 KWNH) is the hottest and coldest temperatures recorded at weather stations provided by reliable networks of weather dataset given a first-level quality control/assured (QC/QA) by a WPC meteorologist for a given calendar day across the 48 contiguous U.S. states and Washington, DC.
Something is different. What changed and why?
On July 22, 2026, the WPC National High/Low was no longer required to be imported into the NWS Selected Cites (SCS04 or FPUS20 KWBN,Pg. 4) which, prior, had necessitated a four-times (4x) per day issuance. This has changed to a two-times (2x) per day issuance, aligning the high and low to be consistently reporting for a specific calendar day for the product and webpage. This will aid archiving of products as well.
Secondly, a dedicated webpage is now updating regularly along with a geographical map displaying the locations of the sites with the hottest/coldest temperatures found. Having a map to find where the observations were taken was a needed improvement.
How frequently is the daily High and Low updated?
Officially twice. The first at 0230Z (930 PM Eastern Standard Time or 1030 PM Eastern Daylight Time), the second 15.5 hours later at 1800Z (100 PM Eastern Standard Time or 200 PM Eastern Daylight Time the next day) to allow for late arriving CO-OP sites (note: only high temperatures) to be evaluated to replace/add, if needed.
Note: The 0230Z and 1800Z issuances are in UTC and, therefore, are issued one calendar day after the actual temperature data is reported. The TUESDAY High and Low would be sent in UTC time on WEDNESDAY (see bullet below).
Was the Daily High and Low observed on that specific calendar day?
Yes and No.
Yes: The Daily High is considered the highest reliable temperature reported in the Lower 48 states, between LOCAL midnight to midnight.
No: The Daily Low is actually considered from LOCAL noon the prior calendar day to noon of the official report date. The vast majority of Low temperatures DO occur early in the morning on the date itself. However, on rare occasions, the Low may have occurred prior to midnight the day before. For example, a Low may occur at 11pm on Tuesday, but the temperature rose afterward. If this is the coldest reliable location, the Low at 11pm would be considered the Low for WEDNESDAY, not Tuesday.
Note: the final issuance for a calendar day is actually posted with the 1800Z issuance the subsequent day; e.g., Jan 1st High/Low was finalized and sent at 180Z on Jan 2nd.
What weather observations are used for determining the Daily High and Low?
Sites used are from data networks or providers deemed reliable with statistically rigorous data quality assurance, including but not limited to sensor sensitivity, placement and timeliness. This may include other government-run agencies with weather observation networks within the Lower 48 states and may incorporate some local, state-run or those considered reliable through peer reviewed/refereed journal articles.
This includes, but not limited to:
- National Weather Service - Automated Surface Observation Systems (NWS ASOS)
- Federal Aviation Administration - Automated Weather Observation System (FAA AWOS)
- National Interagency Fire Center - Remote Automated Weather Stations (NIFC RAWS)
- NWS Cooperative Observer Program (CO-OP)
- NWS Run Mesonets - Example: San Diego (SGX) or Boulder (BOU)
- State/University Funded Mesonets, including OK, NY, West TX, KS, KY, etc...
- Utah State Climate Center - Peter Sinks
- U.S. Climate Reference Network (USCRN or CRN)
Any additional weather stations that are routinely reported in NWS Forecast office Regional Temperature and Precipitation Products (RTP) and or Hydrometeorological Products (HYD), including commonly referred stations such as: Death Valley, CA or Estcourt Station, ME.
This may include non-routine reports that a NWS Forecast Office disseminates via their official social media accounts (Twitter or Facebook).
How is the data vetted? How could a weather observation potentially be missed?
The determination of the daily maximum and minimum temperatures at WPC is NOT fully automated. There are automated methods to compile and visualize the vast amount of Maximum and Minimum temperatures, but there is always a forecaster doing the first level of quality control. However, some data may arrive afterwards, even one or two days later (more likely for NWS CO-OP reports). As such, we may not have all the data and so the results are considered Preliminary. Official records will be kept by the National Center for Environmental Information (NCEI).
I found a colder/hotter weather report for today.
This is quite possible. Here are a few possible reasons:
- The observation is from a weather site that WPC believes may have some data quality concerns.
- It is a carry-over observation. A "carry-over" is typically associated with a 24hr Low temperature that occurred PRIOR to Noon (i.e. the Minimum temperature for the prior calendar day). This may even occur with automated ASOS and AWOS stations. This is quite often the most difficult observation to fully deduce precisely when the Low occurred and may lead to errors.
- It is a late-arriving CO-OP report. For example, a morning observation is received at 8am reporting the weather to NWS forecast office staff with the Max/Min temps, precipitation, etc... for the last 24hrs or even sometimes longer (if the observer missed one or two days).
Why isn't my home, school, or local TV station's data used?
Even though there are weather observations that may be very reliable, well-sited and show high statistical correlation to surrounding reliable networks, there is often too much data to sift to ascertain its reliability. If you wish your data to be considered for National Hi/Lo, consider becoming a NWS Cooperative Observer (CO-OP).
Why is a location with very low or no population used? Why is a mountain peak used for the Low temperature?
In the past, there were requirements for city population and elevation. Yet, this excluded places of interest, such as Death Valley, Mount Washington in New Hampshire, Mount Mitchell in North Carolina, etc... Also, populations can change both over years but also seasonally. Additionally, mountains in the East fell below the prior elevation criteria resulting in unsatisfying and inconsistent application of rules. For simplicity it was determined to remove the requirements that were reducing timeliness and first-order quality assessment of candidate observations.
