Recent attempts to reconcile broadband shortwave irradiance measurements and models suggest that the problems lie with the measurement of diffuse horizontal irradiance. Broadband diffuse irradiance is, perhaps, best measured using a pyranometer and a tracking disk or ball to shade the direct solar beam. Pyranometers that use a thermopile to measure irradiance proportional to the difference in temperature between a heat sink and an absorbing black surface have shown significant negative offsets because the absorbing surface cools by radiating infrared, which is most significant for clear, dry skies.

An intensive observation period of simultaneous diffuse measurements was conducted in early Autumn of 2001 to assess the similarities and differences among 15 independent measurements of diffuse irradiance. These measurements were made by instruments that used correction procedures for the effect described above and by instruments whose contributers deemed corrections unnecessary by virtue of instrument design. The participants provided an absolute calibration of their instruments that that did not use the data taken during the four-week, intensive observation period. The results of these measurements are compared. The near-term goal is to derive a concensus standard for this measurement in lieu of an absolute standard for diffuse horizontal irradiance.