Waveguide Loss is Not a Measureable Quantity

Some of the weather radar equations used as the basis for calibration of reflectivity measurements contain the waveguide loss as an explicit factor. However, in practice the waveguide loss in an installed radar system is not a measureable quantity. The straightforward procedure involves injecting a known CW test signal at one point, near the directional coupler used for transmit power measurements and receiver calibration, and measuring the power received at another point on the antenna. The first problem is that it is rarely, if ever, possible to incorporate the entire waveguide run from the directional coupler to the antenna feed horn in this measurement.

Moreover, the measured loss involves a combination of the intended ohmic losses along the waveguide run and the mismatch losses related to impedance discontinuities along the path. The measured value consequently varies with frequency; the radar transmit signal spans a range of frequencies, and the test setup disrupts the mismatch configuration present in actual operation. In practice the two loss components cannot be separated, and it is therefore unclear what waveguide loss value applies under normal operating conditions. The variation in the loss measurement, while perhaps not large, will often be adequate to cause significant uncertainty in the overall reflectivity calibration (since the loss enters twice into the radar signal path). The problem will be illustrated by data from an actual S-band radar system.

The preferred solution is to work from measurements at the forward port of the directional coupler. Any waveguide losses up to that point are accounted for in the transmit power measurement and the receiver calibration. Losses on the antenna side of the coupler are incorporated into the effective antenna system gain, as measured with a standard-gain horn in the far field or a calibration sphere. Consequently no direct measurement of waveguide loss is needed.