| Draft accuracy, precision, and random error targets for SAMOS. Accuracy estimates are currently based on time scales for climate studies (i.e., 10 W/m2 for Qnet). Several targets are still to be determined. Drafted: Summer 2004. | |||
| Parameter | Accuracy of Mean (bias) | Data Precision | Random Error (uncertainty) |
|---|---|---|---|
| Latitude and Longitude | 0.001° | 0.001° | |
| Heading | 2° | 0.1° | |
| Course over ground | 2° | 0.1° | |
| Speed over ground | Larger of 2% or 0.2 m/s | 0.1 m/s | Greater of 10% or 0.5 m/s |
| Speed over water | Larger of 2% or 0.2 m/s | 0.1 m/s | Greater of 10% or 0.5 m/s |
| Wind direction | 2° | 1° | |
| Wind speed | Larger of 2% or 0.2 m/s | 0.1 m/s | Greater of 10% or 0.5 m/s |
| Atmospheric Pressure | 0.5 hPa | 0.01 hPa | |
| Air Temperature | 0.1°C | 0.05°C | |
| Dewpoint Temperature | 1.5°C | 0.1°C | |
| Wet-bulb Temperature | 1.5°C | 0.1°C | |
| Relative Humidity | 1% | 0.5% | |
| Specific Humidity | 0.15 g/kg | 0.1 g/kg | |
| Precipitation | ~0.4 mm/day | 0.25 mm | |
| Radiation (SW in, LW in) | 2-3 W/m2 | 1 W/m2 | |
| Sea Temperature | 0.1°C | 0.05°C | |
| Salinity | |||
| Surface Current | |||
The target accuracies are based primarily on the those outlined by the Tropical Ocean Global Atmosphere/Coupled Ocean-Atmosphere Response Experiment (TOGA/COARE) and WOCE programs. The targets should be achievable with existing instrument technology as long as the sensors are constantly monitored during a cruise (C. Fairall, personal communication, 2004). Whether they can be achieved by fully autonomous sensors is still an open question. One concern with most accuracy targets for climate applications is that the estimates tend to only include the accuracy of the mean. This may not be sufficient and SAMOS participants suggest that accuracies should also include a target value for random error (not yet fully developed). To better achieve these targets, the SAMOS initiative plans to "translate" these accuracy targets from the goals proposed for climate applications to accuracy targets that are of use to ship technicians and instrument developers.
Other issues related to accuracy are still under discussion in the SAMOS community. First off, accuracy targets for wind direction should vary with the wind speed; however, these varying accuracies for direction cannot be addressed until there is agreement on the overall target accuracies. Another issue focuses on accuracies needed for GPS information (position, speed, course, and in some cases heading). GPS technology is improving much faster than meteorological sensors and can resolve positions to much higher accuracy than 1 km. The 2nd workshop panel agreed that knowing the ships position within a kilometer will be sufficient for most satellite and flux applications, with coarser accuracy being adequate for modelers. Finally, the SAMOS community opened discussion on the accuracy to which the instrument height/depth must be known. Changes in load can significantly alter the height of instruments on merchant vessels, but the consensus was that these changes are generally small (1-2 m) on R/Vs. Several panelist at the 2nd workshop suggested that placing height sensitive instruments as high as possible on a vessel will, to some extent, alleviate this problem on R/Vs. Handling height changes on merchant vessels (sometimes up to 10 m, but more often 2-3 m [Elizabeth Kent, personal communication, 2004]) is still an open question.