Central Iowa Dew Point Comparison-DEWPEX


Kimberly Schafer and Jennifer Stark
National Weather Service Forecast Office
Des Moines, Iowa



During the last few years, the National Weather Service (NWS) has been commissioning Automated Surface Observing Systems, or ASOS sites, across the country. To supplement these sites, the NWS in Des Moines has utilized a mesoscale network implemented by the Iowa Department of Transportation (IDOT) called the Automated Weather Observation System, or AWOS. From the AWOS sites, forecasters are able to receive some of the same weather information that is available through ASOS: visibility, barometric pressure, cloud height, temperature, dew point and wind data. Precipitation is not available at this time. The NWS office in Des Moines uses the AWOS network extensively, especially during the summer months when dew points are higher than 60 degree F. Instability, heat indices and humidity levels are calculated routinely across the state.

When compared subjectively during the summer months, AWOS dew point temperatures seemed higher than ASOS dew point temperatures at larger airport sites. An in depth field experiment at four central Iowa AWOS sites was designed to double check AWOS dew point values versus sling psychrometer values, referred to as Bias (eq. 1).



 Bias Dew Point = AWOS Dew Point - Sling Psychrometer Dew Point


The goal of the experiment was to prove if AWOS dew point temperatures were indeed too high, and to hypothesize what would cause the problem. The following sections will discuss whether the high dew points were a phenomenon of the corn growing season, a sensor calibration problem or the result of vegetation (weeds, beans, etc.) surrounding the site. With the vast corn crop across the state, forecasters began to think that the high dew points were a result of evapotranspiration from the plants and that the higher moisture content was not being sampled at larger airports.


The dew point experiment (DEWPEX) used sling psychrometers to obtain wet-bulb and ambient temperatures for AWOS sites in central Iowa and compare the resulting dew points to those measured by AWOS. The DEWPEX team decided to start the experiment in the pre-canopy stage of the corn crop to determine if indeed corn was increasing the dew points. The four stations selected for the experiment included Ankeny, Boone, Newton and Knoxville (Figure 1). Each of the stations is located within an hour drive of the NWS office in Johnston which allowed easy access for the staff. After a few phone calls to airport managers, access was gained to the AWOS sites.

Figure 1. DEWPEX AWOS sites.

The experiment began by comparing two sling psychrometers with an automated Maximum/Minimum Temperature System (MMTS) at the Johnston NWS office. The thermometers were found to be well within the calibration limits required for AWOS. Field measurements began in June 1996 and continued through the beginning of August 1996. A typical DEWPEX day included verifying that dew points across the state were above 60 degrees, and that there had been no rainfall for the past six hours. The data was retrieved between 1:00 p.m. and 4:00 p.m. CDT when observers were available. Ideally, target sling times would have been from 7:00 a.m. to 9:00 a.m. CDT, when relative humidity levels were high, and between 3:00 p.m. and 5:00 p.m. CDT, when dew point values were high. Each team of observers, consisting of NWS employees, went to the AWOS site for approximately one hour and took temperature and dew point measurements every ten minutes. Observers also completed a site survey including a description of vegetation and its height surrounding the AWOS instruments, significant terrain features such as valleys, wind direction and the location of any nearby corn. Once back at the office, the observers used a modem to call the AWOS site and download data corresponding to the observation times. The sling dew point temperature was calculated from the wet-bulb and ambient temperature data. Throughout the summer, several staff members collected three data sets (dry bulb, dew point, and wet bulb temperatures) at the four AWOS sites making 12 sets of data, which allowed for enough information to start obtaining temperature trends for each site.


The AWOS system was designed similar to ASOS, so the research process began by studying ASOS temperature, dew point and siting requirements. The ambient temperature from ASOS is derived using a hygrothermometer. Temperature is measured every minute and stored for 12 hours. A quality control program is completed with each measurement to check that the temperature is within the allotted range of -80 degree F to 130 degree F and that it is not more than 6 degree F warmer than the previous temperature. If either of the criteria is not met, the program will report the temperature data as missing. The reported temperature is actually a five minute average. This value is obtained by using four of the last five one-minute temperature values and then averaging them for a five minute temperature value. The five minute temperature value is updated every minute and stored in ASOS for 12 hours.

Dew point measurements are taken in a method similar to temperature. The dew point sensor instrumentation has a chilled mirror that condenses moisture on the surface. When the sensor detects condensation, the instrument determines what temperature is needed to keep the sensor at equilibrium with the surrounding air. This temperature is the dew point. A dew point measurement is taken every minute and stored for 12 hours. The quality control program checks the dew point to make sure that the measurement is less than 6 degree F greater than the previous dew point and that it is not two 2 degree F warmer than the ambient temperature recorded at the same time. If either of these criteria are not met, the dew point will be reported as missing. As with the ambient temperature, the dew point reported is a five minute average and stored in ASOS for 12 hours.

Recommended siting criteria for ASOS and AWOS are that sensors be mounted five feet above the ground or two feet above the average maximum snow depth. The sensors should also be installed away from artificial conditions such as large buildings, cooling towers and large expanses of concrete. Any grass or vegetation within 100 feet of the sensor should be less than ten inches in height. All of the sites in the study met the above criteria with the exception of the vegetation recommendation. Two of the sites had three-to-four foot high vegetation growing around the sensors during the study.

AWOS sites are owned by the IDOT. Coordination calls to the equipment contractor confirmed that the AWOS data is obtained through the same methods listed above for ASOS (Fairchild Communications, Boone, IA, personal correspondence). Quarterly and annual inspections of the sites have been completed by the IDOT equipment contractor. However, temperatures are only calibrated annually and not at the quarterly inspection. The main foci of the quarterly inspection are the barometers, wind equipment, visibility sensors and tipping buckets. Temperature sensors are calibrated with a tolerance of 2 degree F for ambient temperature and 2½ degree F for dew point (IDOT, Ames, IA, personal correspondence). It has been noticed by the contractor that when the temperature is near freezing and the spread between ambient and dew point temperature is small, there tends to be a greater error in measurement.


The Ankeny airport was visited four times. On all occasions, the AWOS temperature was too high when compared with the sling psychrometer data, and only one third of the data was within AWOS calibration limits listed above. Dew points, on the other hand, were within calibration limits for three quarters of the data set. The only bad dew point data appeared to be from one day early in June. Figures 2a and 2b show the ambient and dew point temperature bias for all of the data collected at Ankeny. The Ankeny site was one that did not meet siting criteria set by ASOS/AWOS. The sensor is located next to a marsh area where weeds and grass grew up to six feet tall during the summer. East of the sensor, the grass between the runway and the site started out two feet tall in early June, but was kept cut the remainder of the study. The original thought of the corn crop affecting the sensors was not a problem in Ankeny but it appeared that the marsh land might have a large effect on the dew point data. The temperature error seemed to be a sensor calibration problem.

Figure 2. Part A is Ankeny temperature bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation. Part B is Ankeny dew point bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation.

The next site visited was the Boone airport. Once again, temperatures were too high when compared with the sling psychrometer (Figure 3a and 3b). However, for Boone, the data was within the calibration limits two thirds of the time. The main problem at Boone confirmed forecasters initial thinking, the dew point was too high roughly two thirds of the time. This could be another siting problem for the dew point, or a calibration problem regarding the ambient temperature. The instrumentation was located in the middle of a hay field adjacent to the runway complex. Through July, the hay field had not been cut, which could be the main reason for the incorrect dew point data. The closest corn crop to the Boone site was several hundred yards away, so it is not believed to have had a direct effect on the sensors.

Figure 3. Part A is Boone temperature bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation. Part B is Boone dew point bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation.

The one site that met all of the ASOS/AWOS criteria was Newton. The bias for ambient temperature was within limits and the dew point temperatures were within the limits except for three data points (Figure 4a and 4b). The sensor siting was not much of a problem with soy beans growing next to the instruments in one quadrant and cut grass over the remainder. The closest corn fields were several hundred yards away so corn was not a problem at Newton either. However, near the end of July, weeds growing underneath the sensors were reaching two to three feet exceeding the ten-inch recommendation for maximum vegetation height.

Figure 4. Part A is Newton temperature bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation. Part B is Newton dew point bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation.

The last site, Knoxville, was only visited once during the summer. All of the ambient and dew point temperature data points were within the set criteria. Figures 5a and 5b show the temperature and dew point bias for Knoxville. Siting was correct at Knoxville with no corn reported near the instruments. It was recommended that additional measurements be taken at Knoxville before deciding on the correctness of the sensor data.

Figure 5. Part A is Knoxville temperature bias. Bias-AWOS-Sling psychrometer. Each bar represents one ten-minute observation. Part B is Knoxville dew point bias. Bias=AWOS-Sling psychrometer. Each bar represents one ten-minute observation.


At two sites, Ankeny and Boone, tall vegetation surrounding the site seemed to be the main cause of incorrect dew point temperatures. Temperature calibration could also be a problem, since Ankeny, for example, exhibited high temperatures all summer but dew points generally were within calibration limits. The other two sites had appropriate siting, and in turn both temperature data sets were within the set criteria during the comparison. The hypothesis that the Iowa corn crop was creating high dew points has not been proven with this experiment. None of the four Central Iowa sites had corn growing immediately adjacent to the sensors. However, the four sites are typical of Iowa countryside in which there is a mixture of corn, soybeans, grass and trees nearby with little or no effects of urbanization. Results indicate that tall vegetation is probably a concern for meteorologists. It would be helpful during future growing seasons to have a site survey on hand for each of the Iowa AWOS sites. This would allow the forecaster to know the vegetation surrounding the site and if it could produce incorrect ambient and dew point temperature values. Attempts are being made to work with the IDOT contractors on correcting the Ankeny and Boone AWOS temperature data.


DEWPEX Team members included Karl Jungbluth, Will Kubina, Chuck Myers, Erik Pytlak, Kim Schafer and Jennifer Stark. Thank you to Karl Jungbluth for his input and review of the paper.


DOC/NOAA, 1994: Federal Standard for Siting Meteorological Sensors at Airports - User's guide, Government Printing Office (GSA), FCM-S4-1994, 44pp.

____________, 1994: Algorithms for the Automated Surface Observing System (ASOS) - User's guide, Government Printing Office (GSA), Note 94-4, 110pp.

Stark, J., 1996: Maintenance and Calibration of the AWOS Sites. DMX Tech Time Vol. 1, No. 17, NWS Forecast Office, 9607 NW Beaver Drive, Johnston, IA 50131, 2pp.


USA.gov is the U.S. government's official web portal to all federal, state and local government web resources and services.