One goal of this project is to provide ground validation measurements in support of NASA's EOS/Terra spacecraft. In the case of this particular school, we are fortunate to have a nearby aerosol monitoring site that is part of the AERONET global network of ground-based sun photometers.

The figure below shows AERONET data collected during April and May, 2002. These data are shown as blue and red solid lines. Note that even though these lines seem to imply that the AERONET data are continuous, this is not the case. The measurements are pre-programmed and are made automatically during the day at approximately 15-minute intervals, sky conditions permitting.

The square- and diamond-shaped symbols are GLOBE student data. The GLOBE sun photometer measures aerosol optical depth (AOT) at about 505 (green) and 625 (red) nanometers. In general, these AOT values should fall between the AERONET wavelengths of 440 and 670 nm. For the most part, the student data show excellent agreement with AERONET data. The only notable exceptions occur on April 3 and April 25. The earlier measurement, in particular, probably represents a data reporting error. The source of the differences on April 25 is not known.

The AOT values in the figure are somewhat higher than I would expect to see, even over an urban area. Based just on the GLOBE data, I would suspect that some of the measurements are contaminated by the presence of clouds between the sun and the observer. This may, in fact, be the case on April 3 and April 25. Good agreement with most of the AERONET data tends to dispell such suspicions, as the AERONET data are "screened" for possible cloud contamination before they are made available on the Web. However, this does not necessarily mean that all AERONET AOT values are calculated under acceptable sky conditions. Cloud detection algorithms are not perfect. The screening algorithms rely on characteristics of the data themselves, rather than on actual simultaneous observations of sky conditions. When these algorithms are applied in an automated way, without human intervention, then errors in interpretation can, and will, occur. That fact provides the primary justification for using human observers to collect sun photometer data. And, of course, it explains why accurately reporting metadata is an important part of this protocol.

From a scientific perspective, it is reasonable to conclude from these data that GLOBE students can make reliable measurements of AOT with relatively inexpensive handheld devices (about US$100 vs US$25,000 for the AERONET instruments). AERONET instruments require a great deal of maintenance and there are only a few dozen in regular use around the world. Thus, GLOBE students can greatly augment the data provided by a professionally maintained project such as AERONET.

During the next phase of this project, we will be comparing student aerosol measurements against "aerosol retrievals" provided by the MODIS instrument on NASA's EOS/Terra spacecraft. This task is more challenging because it requires that ground measurements be timed to coincide with Terra flights over the observing site, typically in mid to late morning. These kinds of "ground truth" measurements are important because aerosols are not being monitored over most of the earth's surface, even in developed areas such as Europe. As a result, GLOBE schools are in a position to provide high-quality data that will not be otherwise available.