Home' RTCA Documents for Review : DO-220A Change 1 Contents Appendix C
© 2018 RTCA, Inc.
DO-220A Change 1
Based on the above values, the probability of a missed detection of a 0.130 microburst
is plotted vs. radar range in Figure C-6 for three values of weather reflectivity. Figure
C-7 shows the probability of a nuisance detection of a 0.085 microburst vs. range.
Neither of these plots shows the effect of an SNR threshold. The threshold would make
the plot of missed detection go to one and the plot of nuisance detection go to zero at
ranges where the signal level was below the threshold value.
Figure C-8 plots the detection and nuisance alert probability vs. the single-pulse SNR
level. At low SNR, both curves approach an asymptotic value of 0.5 due to the nature
of the calculation (see Figure C-5).
Figure C-9 plots the ratio of the detection probability to the nuisance alert probability
vs. single-pulse SNR. Figure C-10 is a similar plot except that the post-processed SNR
(based on 128 pulses) is used as the ordinate. These curves permit selection of an SNR
threshold that will provide a desirable single scan ratio of detection to nuisance alert.
A value of log 1 for this ratio (i.e., 1 nuisance alert per 10 true alerts) will be attained
using an SNR threshold of approximately -3 dB (single pulse SNR) or + 7.5 dB
(processed SNR). This value is used for the following cumulative probability
Cumulative probability calculation using equations C-14 and C-15 are shown in Figure
C-11, Figure C-12, and Figure C-13 for -5, 0, and +5 dBZ respectively. The detection
probability rapidly goes to unity just after the SNR threshold is reached. The nuisance
alert probabilities reach an asymptotic value as shown.
Probability Calculations with the NASA Clutter Suppression Techniques Considered
The above calculations were made assuming no hazard area threshold or multiple scan
detection was used. Calculations using these features have also been made and are
plotted in Figure C-14 through Figure C-18.
Figure C-14 and Figure C-15 plot the probability of missed detection and nuisance
detection for the NASA system using hazard detection on two radar scans prior to
declaring an alert and a hazard area threshold of 0.2 sq. km. The steps in the plots are
caused by the multiple scan detection requirement and the hazard area thresholding
The associated cumulative probability plots using an SNR threshold of -3 dB (single
pulse) are shown in Figure C-16, Figure C-17, and Figure C-18. Notice that the
probability of a nuisance alert has been reduced considerably and the probability of a
missed detection has been increased somewhat at a given range.
Summary and Conclusions
The equations and calculations discussed above provide a useful exercise in determining
the expected radar system performance as limited by receiver noise. They also can provide
a guide to threshold settings and the effect of these thresholds on detection performance.
The threshold settings used in the calculations have not been optimized, and further
calculations will be made to adjust the thresholds to provide an optimized tradeoff between
detection and nuisance alert performance.
Links Archive DO-356A ED-255_DO-xx for OC Navigation Previous Page Next Page