MONITORING STATION INSTRUMENTATION AND MEASUREMENTS - Wind Resource Assessment: A Practical Guide to Developing a Wind Project - page 45

4.2.2 Heated Anemometers
The build up of ice can cause anemometers to rotate more slowly, stop, or even
break because of the load (e.g., falling ice). Likewise, it can cause wind vanes to be
off-balance or alter their aerodynamic profile, distorting the directional readings, or
can freeze them in one position. For these reasons, where frequent or heavy icing is
expected, met towers should be equipped with at least one or two heated anemometers
and direction vanes. Unfortunately, heated sensors consume much more power than
unheated sensors (power supply options are discussed in Section 4.7).
Except during periods of icing, heated anemometers are generally less accurate
than unheated ones. Therefore, it is recommended that unheated anemometers be
the primary source of wind data, while heated ones should be used only to fill gaps in
the primary data record. In addition, to maintain as much consistency as possible in the
heated-anemometer readings, it is recommended that power be applied to the heating
elements throughout the year, not just in the cold season. In a typical configuration,
an unheated anemometer might be paired with a heated anemometer on each of the
top two levels of the mast, and one of the two direction vanes on the mast might be
heated. This approach strikes a balance between high overall data recovery and good
measurement accuracy.

4.2.3 Delta Temperature
The parameter T (pronounced delta tee) is the difference in temperature between

two heights above ground. It is a measure of the thermal stability, or buoyancy, of
the atmosphere. The challenge is to measure the temperature difference with sufficient
accuracy to be useful. According to the EPA Quality Assurance Handbook (1989),
the maximum allowable

T error is 0.003 C/m (degrees Celsius per meter of height).
With heights of, say, 10 and 40 m, the allowable error is just 0 1 ◦ C. To achieve

this, a pair of identical temperature-sensing subsystems calibrated and matched by
the manufacturer is typically used. To maximize the height difference and thus the
precision of the result, one sensor should be placed about 3 m above the ground and
the other a short distance (e.g., 2 m) below the top sensing level on the tower. In
addition, both sensors need to be mounted and shielded in the same manner so they
respond similarly to ambient conditions. To further reduce errors, a radiation shield
that uses either forced (mechanical) or natural (passive) aspiration is required (to meet
EPA guidelines, forced aspiration may be necessary). The data logger manufacturer
should be consulted to determine compatible sensor types and models.

.

4.2.4 Barometric Pressure
Knowing the barometric pressure along with the air temperature can help improve
the accuracy of air density estimates, as normal variations in pressure at the same
temperature can affect air density by about 1%. However, barometric pressure is dif-
ficult to measure accurately in windy environments because of the dynamic pressures
induced by the wind passing across the instrument enclosure. For this reason, high