2 Chapter 2 — Station Networks: The Foundation of Everything
Everything the IEM does starts with a station. A chart of daily temperature departures needs a station. A warning search needs a county or a Weather Forecast Office (WFO). A precipitation download needs to know which network you want and which observer. Before you can use any of the IEM’s tools effectively, you need to understand how it organizes the raw material — the observing stations — and what each network actually represents.
The IEM ingests data from more than a dozen distinct observing networks, each with its own history, its own measurement standards, its own strengths, and its own quirks. This chapter covers the networks you will actually encounter when doing weather research: what they measure, how far back they go, where their limitations are, and how to navigate them on the IEM.
Finding Any Station
Before diving into specific networks, the most important page to know is the station locator at mesonet.agron.iastate.edu/sites/locate.php. Type a city, a station ID, or a ZIP code, select a network from the dropdown, and the IEM returns a list of matching stations with their coordinates, elevation, archive start date, and a link to that station’s full data page. This is your entry point whenever you need to identify the right station ID before pulling data.
If you want to browse all stations in a given network, the Network Tables page at mesonet.agron.iastate.edu/sites/networks.php lists every network the IEM tracks — from Alabama ASOS to Zimbabwe ASOS, and everything in between. For a given network you can generate the station list as an HTML table, CSV, shapefile, GEMPAK table, or MADIS table. Each station entry links to its individual station page, which shows its attributes — including things like whether it has 1-minute data, whether it has TAF forecasts, what its GHCNH identifier is for cross-referencing with NCEI, and when the archive officially begins.
Station IDs on the IEM follow a few different conventions depending on the network. ASOS stations use 3-character domestic identifiers (DSM for Des Moines, MDT for Harrisburg) or the full 4-character ICAO identifiers with the K prefix (KDSM, KMDT). COOP stations use IEM-internal identifiers like IATDSM or station codes from the GHCND database (USC00######, USW000######). DCP stations use the 5-character NWS Location Identifiers (NWSLI) that appear in SHEF product headers. The station page for any station shows all the relevant identifiers in one place, which avoids a great deal of confusion when cross-referencing data between systems.
ASOS and AWOS: The Airport Backbone
The Automated Surface Observing System — ASOS — is the workhorse of the IEM for weather hobbyists. These are the automated weather stations at airports across the United States, operated jointly by the NWS, the FAA, and the Department of Defense. They report continuously, have long records, and measure the variables most users care about: temperature, dew point, wind speed and direction, visibility, ceiling, precipitation, and pressure.
ASOS stations record temperature in whole degrees Fahrenheit internally, but transmit via the METAR format, which requires Celsius. The conversion introduces a rounding artifact: 78°F becomes 25.56°C, transmits as 26°C, then converts back to 78.8°F — which rounds to 79°F. A full degree of phantom warming can enter the record purely from the transmission pipeline.
If you are doing precise temperature analysis, especially around record extremes, look for the supplemental T-group in the raw METAR feed; it carries the unrounded value. This is covered in depth in Chapter 10.
The IEM’s ASOS mainpage is at mesonet.agron.iastate.edu/ASOS/. From there you can get to current conditions, recent METAR reports, monthly precipitation tables, download portals, and the 1-minute archive.
A few points are worth keeping in mind before you rely heavily on ASOS data:
Archive depth varies considerably by station. The major airport stations — places like Des Moines (DSM, January 1928), Burlington (BRL, September 1931), Mason City (MCW, October 1940) — have records stretching back decades before ASOS hardware was even invented. That is because those stations were manually observed airports before being converted to automation. Smaller stations came online in the mid-1990s when ASOS was deployed nationally; most of those have archives starting between 1994 and 1996. When you pull ASOS data for a station, always check the archive begin date in the station metadata before assuming the record goes back as far as you need.
ASOS is not AWOS, and the distinction matters. The Automated Weather Observing System (AWOS) is a slightly older, less capable automated system found at smaller airports. AWOS stations generally report less frequently than ASOS, do not always report precipitation, and have less rigorous quality control. The IEM lumps ASOS and AWOS together under the same network category, but when you look at a station’s attributes you will see IS_AWOS=1 on AWOS stations. For serious precipitation analysis, you want ASOS stations — specifically the ones with HAS_PHOUR=1 (hourly precip bucket) or HAS1MIN=1 (1-minute archive). Those attributes appear in the station table and on each station’s page.
Temperature is stored internally in whole-degree Fahrenheit. This is a design limitation of the original ASOS sensors. For most purposes it makes no difference, but if you are doing analysis that requires sub-degree temperature precision — onset of freezing at a specific threshold, for example — use the T-group from the raw METAR string when it is available. The IEM’s hourly download includes the raw METAR field, so you can parse the T-group yourself.
The precipitation record has a well-known gap around the ASOS transition. When ASOS replaced manual observers at airports in the early-to-mid 1990s, precipitation measurement changed. The original ASOS tipping-bucket gauges had documented problems with snowfall and light precipitation. A program to install All-Weather Precipitation Accumulation Gauges (AWPAG) at ASOS sites addressed some of these issues, but the install dates varied by station and the improvement was gradual. For precipitation analysis at ASOS stations that spans the pre- and post-automation period, treat the data on either side of that transition with awareness that the measurement method changed.
The 1-minute ASOS archive is one of the IEM’s more useful hidden features. The National Center for Environmental Information (NCEI) maintains a 1-minute resolution dataset for many US ASOS sites going back to approximately 2000. The IEM processes this archive daily — it is not real-time, with a delay of 18–36 hours — and makes it available for download at mesonet.agron.iastate.edu/request/asos/1min.phtml and for plotting via Autoplot chart q=211. This is the right data source when you want to look at the precise timing of a temperature drop, a pressure jump, or a precipitation onset during an event. Not every station has it, and the coverage gets thinner the further back you go, but for the past ten to fifteen years at most major ASOS sites it is quite complete.
NWS COOP: A Century of Daily Observations
The NWS Cooperative Observer Program — universally called COOP — is where the IEM’s real historical depth lives. These are the cooperative observer stations: human volunteers, mostly private citizens, who take daily readings of temperature and precipitation. They read a max-min thermometer once a day, measure whatever fell in the rain gauge, and call or submit it to the NWS. Some stations have been doing this continuously since the 1880s.
That longevity is exactly why the data has quirks that any serious user needs to understand.
The IEM’s COOP mainpage is at mesonet.agron.iastate.edu/COOP/, and it is one of the more feature-rich network pages on the site. It gives you access to current observations, the full historical download, the normals download, freezing date statistics, snow depth duration tools, the growing season plotter, and the daily extremes table. But before using any of those, understand what the raw data actually is.
COOP reports are once-daily observations, not instantaneous readings. An observer reads the maximum and minimum thermometer at a fixed time each day — traditionally 7 AM, though some observe at other times — resets the thermometer, and records whatever accumulated in the gauge since the previous reading. This means a “daily” high temperature reading is the maximum over a roughly 24-hour period ending at the observation time, and a “daily” precipitation reading is the accumulation over that same period. This is fundamentally different from the way airport ASOS stations measure daily summaries, where the day runs midnight to midnight.
This observation time offset matters enormously when comparing COOP data to ASOS data, when looking at events that span observation times, and when computing things like monthly totals that might not add up the way you expect. The IEM processes COOP data into its Climodat dataset, which applies quality control and attempts to handle some of these timing issues, but you should be aware that the raw COOP and the processed Climodat are two different things. The Climodat download is at mesonet.agron.iastate.edu/request/coop/fe.phtml (quality-controlled) and the raw COOP observations are at mesonet.agron.iastate.edu/request/coop/obs-fe.phtml.
The COOP network is also where you get access to the IEM’s daily normals and records — the period-of-record daily statistics that underpin most of the climatology tools. The extremes table at mesonet.agron.iastate.edu/COOP/extremes.php shows daily records going back as far as the station record allows, all labeled as unofficial in the IEM’s standard convention. The normals download at mesonet.agron.iastate.edu/COOP/dl/normals.phtml gives you the full 366-day normal climatology for any station, including average high, average low, average precipitation, record values, growing degree days, and solar radiation.
Two COOP sub-tools deserve special mention. The freeze date tool at mesonet.agron.iastate.edu/COOP/freezing.php gives you the statistics of first and last freeze dates across stations — a question that comes up constantly in discussions of growing seasons, late-season snow events, and climate trends. The snow depth duration tool at mesonet.agron.iastate.edu/COOP/snowd_duration.phtml shows, for a given date, how long snow cover typically persists — useful context when you are watching an early season snowfall and wondering whether it is likely to stick around.
One thing to watch for in COOP data: gaps are common and are not always flagged. Observers miss days. Stations go inactive for months or years and then restart. Equipment changes. The longer the record, the more of this you will encounter. When you pull COOP data for a long period and compute statistics, always check the count of valid observations — the IEM reports this in the normals table as the “years” field — before drawing conclusions from stations with thin records in certain time periods.
CoCoRaHS: The Precipitation Dense Network
The Community Collaborative Rain, Hail and Snow Network is the largest precipitation observing network in the United States by station count, and it is built almost entirely on volunteers with a standard gauge, a training module, and an internet connection. CoCoRaHS observers report daily precipitation, snow, and hail at a much finer spatial density than either ASOS or COOP — a typical metro area might have dozens of CoCoRaHS stations compared to one or two ASOS airports and a handful of COOP sites.
The IEM’s CoCoRaHS page is at mesonet.agron.iastate.edu/cocorahs/, with current observations at mesonet.agron.iastate.edu/cocorahs/current.phtml and an observation listing at mesonet.agron.iastate.edu/cocorahs/obs.phtml.
For precipitation research, CoCoRaHS data is most useful when you want to map the spatial distribution of a specific event — how much fell where during a particular storm, whether the radar QPE estimate was close to what gauges measured, how much a lake effect band produced along its axis versus a few miles off it. The IEM integrates CoCoRaHS into its precipitation maps and has it available in the daily summary download portal alongside ASOS and COOP data.
The limitation is consistency. CoCoRaHS observer density varies dramatically by region, and individual observers join and leave the network constantly. A station with a 2012 start date might have had one observer from 2012 to 2018 and a completely different person since then. This is fine for event-specific analysis but makes CoCoRaHS difficult to use for long-term trend work.
DCP/HADS: Rivers, Reservoirs, and Remote Gauges
The GOES Data Collection Platforms — DCP — are a diverse collection of automated sensors that transmit data via GOES satellite. On the IEM, the DCP section at mesonet.agron.iastate.edu/DCP/ covers the Hydrometeorological Automated Data System (HADS) network, which is essentially the NWS’s aggregation of river gauges, reservoir sensors, snow depth sensors, and other remote automated observations from hundreds of different operating agencies.
The data from DCP stations is encoded in the Standard Hydrometeorological Exchange Format — SHEF — a coding system that assigns two-letter physical element codes to each variable. River stage is HP. Streamflow is QR. Precipitation cumulative is PC. Air temperature is TA. When the IEM displays DCP data, it shows these SHEF codes alongside the values. If you are looking at a river gauge and wondering what HP means, that is your answer: height primary, the stage in feet or meters above a local datum.
The most useful DCP tool for weather hobbyists is the interactive current map at mesonet.agron.iastate.edu/DCP/map.php, which lets you select a SHEF code and see the current value at every DCP station that reports that variable. Selecting HP gives you a map of current river stage across the network — a quick way to assess which rivers are elevated in your region during or after a heavy rain event.
The DCP section also has a “tomb” page at mesonet.agron.iastate.edu/DCP/tomb.phtml — a list of NWSLI identifiers that the IEM has received data for but cannot identify because they do not appear in the NWS Location Identifier database. This is a data quality artifact, not a crisis, but it is the kind of honest transparency about data limitations that characterizes the IEM throughout.
RWIS: Roadway Weather and Pavement Sensors
Roadway Weather Information Systems (RWIS) are operated by state departments of transportation to support winter road maintenance decisions. They measure not just standard atmospheric variables but pavement-specific sensors: pavement surface temperature, subsurface temperature, pavement condition (dry, wet, ice, snow), and sometimes chemical sensors for detecting ice-melting agents.
The IEM’s RWIS mainpage is at mesonet.agron.iastate.edu/RWIS/, with current observations at mesonet.agron.iastate.edu/RWIS/current.phtml and road camera feeds at mesonet.agron.iastate.edu/RWIS/camera.phtml.
For weather enthusiasts, RWIS is most useful in two situations. First, during winter weather events, RWIS pavement temperature data tells you something that no atmospheric sensor can: whether the road surface is actually at or below freezing, which is what determines whether precipitation falls as rain or accumulates as ice. A 34°F air temperature with a 29°F pavement is a very different scenario from 34°F air over a 36°F pavement. Second, RWIS stations often fill geographic gaps in the ASOS network — they are typically sited on highways in areas between airports, so they can give you surface temperature and wind readings in locations that would otherwise have no automated coverage.
The limitation is that RWIS stations were designed for road maintenance, not climatological research. Quality control is less rigorous than ASOS, sensors drift, and stations are sometimes moved when highway construction changes the relevant road segment. Use RWIS for current conditions and event analysis; do not build a long-term climatology from it.
Specialty Networks
Several other networks appear in the IEM’s menu and are worth knowing about even if you use them less frequently.
ISU Soil Moisture Network at mesonet.agron.iastate.edu/agclimate/ is an Iowa-specific agricultural weather network operated by Iowa State University. Beyond the obvious agricultural applications, the soil moisture and temperature data from this network is genuinely interesting during transition-season events — it shows you whether soils are frozen or saturated, which affects how precipitation moves into rivers and whether spring warmth is penetrating the ground. The soil temperature maps that appear on the IEM homepage are generated from this network.
SCAN (Soil Climate Analysis Network) at mesonet.agron.iastate.edu/scan/ is the USDA NRCS equivalent — a national network of automated stations at agricultural research sites measuring soil temperature and moisture at multiple depths, along with standard atmospheric variables. Coverage is thinner than ASOS but the data quality is high.
USCRN (US Climate Reference Network) at mesonet.agron.iastate.edu/uscrn/ is NOAA’s gold-standard reference network, designed specifically to detect long-term climate trends without the site-change and urbanization issues that affect longer ASOS and COOP records. USCRN stations are in pristine, undisturbed locations, triple-redundant, and meticulously maintained. The network has been operational since the early 2000s and the data quality is exceptional. The downside is coverage — there are only about 130 stations across the contiguous US, so your nearest USCRN site may be 100 miles away.
CoCoRaHS was covered above. NLAE Flux at mesonet.agron.iastate.edu/nstl_flux/ covers flux tower data from the National Laboratory for Agriculture and the Environment — specialized eddy covariance measurements of energy and carbon exchange that are not likely to be part of most weather hobbyists’ workflows. The Other network page at mesonet.agron.iastate.edu/other/ aggregates miscellaneous networks including TV station weather networks (KCCI SchoolNet, KELO WeatherNet, KIMT StormNet) and specialized institutional networks.
Station Metadata and the Network Table
Every station on the IEM has a page that shows its full metadata. You reach it either by clicking the station ID in a network table or by going to mesonet.agron.iastate.edu/sites/locate.php and searching. The station page shows coordinates, elevation, archive dates, network membership, and a set of attribute flags that tell you what data is available:
HAS1MIN=1— 1-minute ASOS data availableHASTAF=1— Terminal Aerodrome Forecasts (TAF) issued for this stationHAS_PHOUR=1— hourly precipitation bucket data availableIS_AWOS=1— this is an AWOS station, not ASOSGHCNH_ID— the GHCN-Hourly identifier for cross-referencing with NCEIFLOOR— the earliest reliable date for data from this station (overrides archive begin in some cases)SHEF_6HR_SRC— which DCP gauge is used to supplement 6-hour precip data
These attributes are not just metadata — they tell you what you can actually do with a station. Before building a long analysis pipeline around a specific ASOS station, pull its station page and confirm it has the coverage you need. The FLOOR attribute in particular is worth checking: some stations have archive dates that go back further than the data is actually reliable, and the FLOOR reflects the IEM’s judgment about where the trustworthy record begins.
The Network Tables page also offers direct export of any network’s station list as a CSV or shapefile, which is useful when you want to do spatial analysis on a full network rather than looking up stations one at a time.