Optimizing Thresholds
When defining thresholds for your system, you must experiment to obtain the best combinations for your application.
| Rule | Tradeoffs |
|---|---|
|
Use large numbers of samples (>50). The more samples, the better the estimates of all parameters. |
The antenna must scan more slowly when the number of samples is larger and a slow scan rate may be inconsistent with operational objectives. A scan rate of 1 ... 3 RPM is typical for good estimation, but the slower the better. |
|
Use a high PRF. Doppler estimates tend to be better when the PRF is high. |
The larger the PRF, the shorter the unambiguous range. |
|
Do not use a clutter filter that is broader than necessary to remove the ground clutter. The required filter width depends on the selected scan rate. Use the Ascope utility to observe the width of Doppler spectra at close range to estimate the required width of the clutter filter. To simulate IRIS operation, use the Antenna utility to scan the antenna while the Ascope runs. See IRIS and RDA Utilities Guide (M211316EN). |
A clutter filter that is too broad can damage weather information without improving the clutter cancellation. |
To find the right threshold values for your system, check data from different weather situations (clutter only, rain, snow, convective or widespread precipitation) and verify the situations from independent sources of weather information. Consider making special monitoring products such as low elevation PPIs with color scale down to -20 dBZ, and RAINN products over a long period.
The following table suggests some starting points.
| Parameter | Threshold Starting Point |
|---|---|
| T | LOG |
| Z | LOG & CSR |
| V | SQI & CSR |
| W | SIG & SQI & LOG |
| Dual Pol | LOG |
| LOG LEVEL | 0.75 |
| SQI LEVEL | 0.4 |
| CSR LEVEL | 18.0 |
| SIG LEVEL | 5.0 |
| PMI LEVEL | 0.45 |
| Speckle filters | on |
| If you have this problem... | … first try to change this threshold |
|---|---|
| Mountains | Filter Dop bigger or CSR smaller |
| Underestimation of rain over mountains | CSR bigger |
| Doppler snake (no echo at side wind) | Filter Dop smaller |
| Noise at all elevations | LOG bigger |
| Dots at low elevations | Speckle filter on |
| Second trip echoes | SQI bigger |
| Vanishing strong showers | SQI smaller |