The purpose of OASIS98 was to examine the accuracy of the flux measurements made using OASIS instrumentation. To this end, flux estimates obtained from the ISFF instrumentation were compared to those obtained using OASIS instrumentation. This section will summarize these findings. Data from 30 July 1998 will be emphasized because it was a clear day with all instrumentation operating correctly. All data presented here are one hour running means of five minute averaged data unless stated otherwise.

Details of the measurement techniques and methodology can be found in Brotzge et al., 1999, and Richardson et al., 1999.

1.Net Radiation

The top panel of Fig. 2 is the net radiation on 30 July 1998 for the ISFF and OASIS Eppley systems and the OASIS Kipp & Zonen CNR1 and NR-Lite; the lower panel of Fig. 2 is the net radiation difference relative to the ISFF Eppley system. The Eppley long wave (PIR) data was corrected using the dome and case temperatures. The agreement between all systems can be seen to be approximately ± 30 W m-2. The NR-Lite was adjusted by a "calibration" constant determined from a field experiment (see Brotzge et al., 1999b for a complete discussion). A wind speed correction has not been applied to the NR-Lite data. The total available energy is approximately 600 W m-2.

2.Eddy Correlation Sensible and Latent Heat Fluxes and Ground Heat Flux

Figure 3 shows the ISFF and OASIS ground heat flux and sensible heat and latent heat fluxes calculated using the sonic anemometers and Krypton hygrometers for 30 July 1998; also plotted in Fig. 3 is the flux imbalance using the ISFF Eppley net radiation measurement (Fig. 2). Sensible heat flux was calculated using the sonic temperature that was not corrected for moisture. Approximately 300 W m-2 goes into the sensible heat flux while the remaining 300 W m-2 is split about evenly between latent and ground heat flux. ISFF (OASIS) ground heat flux was an average of three (two) heat flux plates at 5 cm and includes the ground heat storage through the use of the integrating soil PRTs; soil moisture was assumed constant. At night, the magnitude of the flux imbalance is less than 10 W m-2 (20 W m-2) for ISFF (OASIS). During the day, the magnitude of the flux imbalance is less than 90 W m-2 (70 W m-2) for NCAR ISFF (OASIS). The 24 hour average imbalance is Ð10 W m-2 (2 W m-2) for ISFF (OASIS).

The difference between the ISFF and OASIS eddy correlation sensible heat flux was examined for the entire month of July, 1998. In general, the measured fluxes were within 5-10 W m-2 at night and within 40-60 W m-2 during the day. The differences appeared random.

3.Profile Measurements of Sensible Heat

The top panel in Fig. 4 shows the difference between sensible heat flux calculated using the profile technique and the ISFF eddy correlation sensible heat flux for 30 July 1998. The difference is generally less than 20 W m-2 at night and 50-100 W m-2 during the day. Wind speeds for this period were about 3 m s-1 at night and 4 m s-1 during the day. These results are fairly typical with the profile sensible heat flux estimates generally within 50-75 W m-2 of the eddy correlation measurements. The profile technique uses wind speed and air temperature at two levels and is therefore sensitive to errors in either measurement (Brotzge and Crawford, 1999).

The profile sensible heat fluxes appear reasonably accurate in light of the differences seen between the two eddy correlation sensible heat flux measurements discussed in section 3.2.

4.Residual Latent Heat Flux

The bottom panel in Fig. 4 shows the difference between the latent heat flux estimated as a residual from the energy balance equation and the latent heat flux measured using the eddy correlation technique. Differences are generally less than 50 W m-2. Net radiation measured by an NR-Lite and the sensible heat flux calculated using the profile technique was used; therefore, this is a reasonable estimate of the error to be expected from OASIS measurements under conditions similar to that experienced during the OASIS98 field project. The results shown are fairly typical, i.e., the residual latent heat flux estimates were generally within 50-75 W m-2 of the eddy correlation measurements.