Ground Penetrating Radar Performance
How IDS is continuously improving ground penetrating radar performance
Ground penetrating radar (GPR) performance is primarily governed by the material being surveyed as radio waves decrease exponentially and soon become undetectable in energy absorbing materials such as wet clay.
This is a physical limit and no amount of instrumentation upgrading will overcome it. However, development efforts aimed at improving the over-all sensitivity of the system may enhance performance in several circumstances and allow even better results.
Since IDS is continuously focusing on improving its Ground Penetrating Radar products, this article analyses some features of ground penetrating radar and highlights their impact on the overall system’s performance. These are:
- Dynamic range;
- Signal stability;
- System internal clutter profile.
This is a common evaluation parameter for a specific GPR system. It represents the maximum total attenuation loss during the two-way transit of the electromagnetic wave which will still allow reception; greater losses will not be recorded.
For instance, in a soil with a two-way attenuation of 40 dB/m, a GPR system with 100 dB of dynamic range will collect a useful signal up to a depth of 2.5 m; beyond such a depth, echoes from targets will become undetectable.
Given that the ground penetrating radar user has no control over the soil attenuation, it is obvious that the dynamic range directly influences the exploration depth of the GPR; as the larger the dynamic range, the greater the GPR penetration.1
The dynamic range can also be defined as the ratio between maximum receivable signal and minimum detectable signal. Great care is therefore needed in GPR design to reduce sources of internal noise and to shield the antenna against external noise as deeper targets may become undetectable if the noise level is too high (see Fig. 1).
Fig. 1: Comparison between an IDS K2 system connected to a 200 MHz antenna (left) and a competitor’s GPR (right) using a 250 MHz antenna; the higher noise level in the latter hides any target deeper than 2m.
It is also possible to improve the dynamic range by averaging (stacking) the received signal (see Fig. 2). The idea is that each individual scan consists of the same signal, except for the noise. Thus, by stacking several traces collected from the same position, contributions from the target will add, whereas the random noise will tend to reduce. The expected improvement in noise reduction is related to the averaging factor.
However, averaging traces impacts on the speed at which the ground penetrating radar is moved, because, unless the physical location is sensibly constant during averaging, data will become contaminated by adjacent samples. In other words, stacking requires a longer collection time and there is an upper limit due to the operational constraints (i.e. the operator requires a minimum value for the scan velocity).
For this reason pulse repetition frequency is a key parameter for a GPR as the shorter the interval between two consecutive pulses emitted by the GPR, the higher the GPR’s maximum speed and consequently, a bigger stacking factor is allowed.
IDS’s K2 control unit features a pulse repetition frequency of 400 KHz, four times greater than competitors. Consequently, a higher stacking factor can be used without affecting the scan velocity and it also helps to increase the dynamic range.
Fig. 2: Improvement of exploration depth due to stacking. The target marked with an arrow in the picture on the left is hidden by noise in the picture on the right.
It has been stated that averaging has an effect on random noise only, meaning that contributions from external sources of noise (e.g. broadcasting aerials, cell phones, etc.) can be reduced by stacking, whereas spurious signals due to a system’s internal reflections cannot be mitigated because they add in the same way as useful echoes from targets. Consequently, ground penetrating radar internal clutter needs to be minimized with a suitable equipment design. This matter will be further analyzed in a dedicated section later in this article.
For evaluating a GPR’s internal noise (i.e. the one generated within the system), a very easy test can be performed. It requires some traces to be recorded with hardware data processing disabled and the antenna disconnected from the control unit.
Just by plotting the averaged power of the stored traces, it is possible to have an idea of the noise level of the equipment under test.
Fig. 3 shows a comparison between an IDS K2 system and a competitor’s ground penetrating radar. As a result of its digital technology, the internal noise power level of the IDS equipment is about 10 times lower (10 dB), thus allowing a greater penetration depth.
Fig. 3: Comparison between the internal noise power level of a competitor’s GPR control unit (blue curve) and an IDS RIS K2 unit (red curve). Due to its digital technology the internal noise level of the IDS RIS K2 unit is about 12 dB lower.
Signal Stability (jitter)
Another key aspect the user should consider when evaluating a GPR, is the stability of generated signals or jitter; for the purposes of this article, jitter here refers to the variation in magnitude of a quantity, from trace to trace, when the system is operating under static conditions. In such circumstance, all recorded traces should be similar (ideally equal). If not, traces do not contain the same information even if collected from the same position and stacking can’t be effective for improving penetration depth. This basic concept is explained in the images below. If equal traces are stacked, the resulting waveform has the same shape (Fig. 4). Alternatively, poor system stability may cause distortion in the resulting trace (Fig. 5).
Fig. 4: If the equal traces (shown on the left) are added, the resulting stacked trace (on the right) has no distortion
Fig. 5: Stacking in an unstable GPR is not effective. The stacked trace (green) is widely different from the ideal waveform (blue).
Another effect of instability concerns data processing. Specifically, some algorithms such as the background removal filter are not able to remove the contribution provided by features that are constant along the whole scan from the jittered data; for instance, it is not possible to completely delete the signal generated by the coupling between the transmitter and the receiver (direct wave) that may mask reflections from important objects just below the subsurface.
Jittering can be related to the amplitude and to the arrival time of the traces and evaluating each contribution is difficult; however, for getting an overall estimation about the stability of a GPR, some signals can be recorded with the antenna in a fixed position. Then, the averaged power of the whole scan can be plotted against the power computed after removing the averaged trace; it can be demonstrated that the difference between the two graphs is a measure of the stability of the system (as the larger the difference, the more stable the GPR).
Fig. 6 shows a comparison between a RIS K2 and a competitor GPR; it can be noted that the K2 shows a stability 18 times higher (25 dB). This noticeable feature is due to the digital technology implemented in the K2 that has increased the stability of the main potential sources of jitter, such as the timing circuit and the analogue to digital converter.
Fig. 6: Comparison in terms of signal stability between a competitor’s GPR system (above) and an IDS RIS K2. Due to its digital technology the K2 signal stability is about 25 dB higher.
System internal clutter profile
Finally, another crucial parameter to be evaluated is the system’s clutter profile.
Clutter is the term used for returns identified by the system as targets that do not correspond to intended targets or to noise; in fact, when a GPR control unit is connected to practical antennas, internal system reflections, ground returns and transmitter to receiver antenna leakage generate time dependent clutter that may hide signals backscattered by the targets.
In this respect, although a large dynamic range is an obvious requirement for an effective GPR, it is not necessarily the major performance limiter. In some equipment it is the case that performance is most often limited by clutter internally generated within the radar system itself.
For that reason, a good ground penetrating radar design also concentrates on the suppression of clutter generation (from internal reflections and from interactions between the antenna and the ground) which sometimes leads to greater improvements in performance than only extending the system’s dynamic range.
Fig. 7 shows a comparison between an IDS RIS K2 system running a 600 MHz antenna and a competitor’s equipment running a 500 MHz antenna. It can be clearly seen that in the radargram on the right, target E is completely hidden by the internal reflections of the system (black and white horizontal stripes).
Fig. 7: Comparison in terms of system clutter between a competitor’s GPR system running a 500 MHz antenna (on the right) and an IDS RIS K2 connected to a 600 MHz antenna (on the left). The horizontal stripes in the competitor’s profile fully hide target E.
1One can think that it is possible to increase the exploration depth just by increasing the transmitting power. Unfortunately, power must increase exponentially in order to increase exploration depth, but there are mandatory limits due to legislation on Electromagnetic Compatibility.