This article was originally published in the September/October 1999 issue of Home Energy Magazine. Some formatting inconsistencies may be evident in older archive content.
| Back to Contents Page | Home Energy Index | About Home Energy |
Home Energy Magazine Online September/October 1999
Weather Data Getting CloudyIn the last two decades, researchers have relied increasingly on computer programs such as DOE-2 and BLAST to analyze building energy use. For example, simulations can be used to extrapolate building energy performance from limited measured energy data, or to weather normalize predicted savings in energy management or shared savings contracts. In order to do a computer simulation, however, one needs detailed, preferably hourly, weather information. The major sources of this weather information are the surface weather observations reported by the National Weather Service (NWS) or the Federal Aviation Administration (FAA) at airports and other sites and archived at the National Climatic Data Center (NCDC). Unfortunately, beginning in 1992, the NWS and the FAA have been changing the method that is used to collect these observations and the types of data that are collected. Credible building energy simulations can no longer be completed using the weather data currently being reported from these sites.
For building energy simulations, the following weather data are needed at a minimum: dry-bulb temperature, some measure of humidity (either wet-bulb or dew point temperature), station pressure, wind direction and speed, and some measure of solar radiation. All of these parameters except the last are generally available in the surface observation data. If measured solar radiation is unavailable, as is the case in most instances, researchers use analytical and empirical models to estimate the amount of total, direct, and diffuse solar radiation from reported conditions of cloud type, cloud cover, and sky cover.
In 1992, the NWS, the FAA, and the Department of Defense began replacing manual observations with automated sensors and reporting systems called variously Automated Surface Observing System (ASOS) or Automated Weather Observation System (AWOS). As of February 1999, there are 694 commissioned NWS and FAA ASOS/AWOS stations in the United States. When the planned 900-plus nationwide NWS/FAA stations are all in place at the end of this year, they will be the primary source of surface observations in the United States.
The ASOS instrumentation was designed primarily to meet aviation needs; data recorded include ambient and dew point temperatures, wind conditions, precipitation accumulation, cloud height to 12,000 ft, and other weather phenomena such as lightning or freezing rain. A laser beam ceilometer provides the only information on sky conditions. In addition to reporting ceiling heights to 12,000 ft, ASOS uses an algorithm that analyzes 80 ceilometer readings over an hour to deduce five levels of sky cover: Clear, Few, Scattered, Broken, or Overcast. Data on cloud cover above 12,000 ft and minutes of sunshine may be included in the future through additional instrumentation or manual augmentation.
ASOS data are reported on a one-minute interval in the Aviation Routine Weather Report (METAR) format. (Information on ASOS can be found on-line at www.nws.noaa.gov/asos.) Hourly data for previous periods of record can be obtained from the NCDC.
The reported ASOS data present major difficulties for those of us who perform building energy calculations because of the very limited information on solar conditions. Furthermore, the automated methods used to determine cloud height and sky cover are incompatible with previously developed solar models. There is a plan to upgrade the ASOS/AWOS network to record minutes of sunshine. But even if this plan is implemented, a new solar model would have to be developed to make such data usable for building energy simulations.
For the present, procedures are urgently needed to estimate solar radiation from the weather elements now being reported by the automated weather stations. For predicting average year energy use, the older weather data are adequate. But for running weather normalization calculations, current data are necessary. (If all you need is average daily temperature data, you may be able to find that information on a Web site maintained by the University of Dayton, where current data, updated daily, is available for 159 U.S. cities--www.engr.udayton.edu/weather.) And only with the development of new procedures can the currently available data be made usable for this type of building energy simulation.
Joe Huang is a staff scientist in the Building Technologies Department at Lawrence Berkeley National Laboratory.
Home Energy can be reached at: email@example.com
- FIRST PAGE
- PREVIOUS PAGE