It is essential to calibrate the echosounder gain if you intend to (1) compare data from a specific site over extended periods or (2) compare data from multiple instruments at a given location. However, it is important to note that echosounder calibration gain serves to quantify transducer loss and may not be necessary for all applications. Currently, the responsibility for calibrating the instrument lies with the user. By lowering a calibration goal (reference sphere) with known echo force down into the sonar beam and placing it in the middle of the beam (acoustic axis), the measured echo force can be compared with the known echo force, just the same way as one can calibrate a weight. For a complete guide on Echosounder calibration, please refer Demer et al. (2015).
The Signature does not have a calibration routine for the echosounder mode. At the moment, the procedure for those who are interested in calibration will be to include calibration values together with the measurements and do the mathematics as a post processing step. Put in another way; the Signature echosounder does not use any of these values in its processing, but they provide users a way to store their calibration values together with the measurements. These only serve as an option for storing calibration data in the instrument which are then output in the data file header.
The calibration values can be saved as polynomials, where the polynomials can be enabled or disabled as needed. The polynomials are presented in the Integrators Guide - Signature together with the commands that are needed to enable them, and it is recommended to take a look in this manual to get an understanding about how to use the commands.
The intention is to calculate the polynomial over the frequency range of the transducer where you use the center frequency from the beam list as origin.
| \( S_V = 10 \log_{10}\left(10^{\frac{P_r}{10}} - 10^{\frac{N_t}{10}}\right) + 20 \log_{10}(R)\) \(+ 2 \alpha R - PL + G_{cal} - 10 \log_{10}\left(\frac{c \tau}{2\,\mathrm{BIN}F}\right) - \psi\) |
(1) | |||
| \( TS = 10 \log_{10}\!\left(10^{\frac{P_r}{10}} - 10^{\frac{N_t}{10}}\right) + 40 \log_{10}(R) + 2 \alpha R - PL + G_{cal} \) |
(2) | |||
| \(S_V\) | - | volume backscattering strength (\( dB\) \( re\) \( 1 m^{-1}\)) | ||
| \(TS\) | - | point backscattering strength data (\( dB\) \( re\) \( 1 m^{2}\)) (Target Strength) | ||
| \(Pr\) | - | \(0.01 dB \cdot i\) where \(i\) is the signed integer value read from the data file | ||
| \(Nt\) | - | noise threshold | ||
| \(c\) | - | speed of sound | ||
| \(G_{cal}\) | - | calculated over frequency using the values in the polynomials (ref:GETUSERECHOGRAM in Intgerator's Guide) | ||
| \(\Psi\) | - | two way beam angle (\( dB\) \( re\) \(1 Steradian\)) calculated over frequency using the values in the polynomials (ref:GETUSERECHOGRAM in Intgerator's Guide and this FAQ) | ||
| \(PL\) | - | configured transmit power level, stored in file header section (ref:GETECHO,PLx where x is the echogram number) | ||
| \(\tau\) | - | configured transmit pulse length, stored in file header section (ref:GETECHO,XMITx where x is the echogram number) | ||
| \(NBINF\) | - | number of frequency bins in echogram | ||
| \(\alpha\) | - | absorption | ||
| \( R\) | - | calculated in instrument using sample timings and nominal sound velocity of 1500 m/s, see Corrections above | ||
The Target Strength of your calibration sphere can be calculated in different ways. We recommend the Standard Sphere Target Strength from AST/NOAA. With a known \(TS\), one can use equation (2) to calculate \(G_{cal}\). By moving the sphere around in the main lobe of the transducer in the far field region, the values above some percentile (90-95\(\%\)) are used as the measured value. To calculate \(S_V\) one needs a number for the two-way beam angle also; we rely on the theoretical value for the two way beam angle since it is difficult to calibrate it unless you have a split beam echosounder. The distance, \(R\), can be found precisely through the pulse compressed echogram through the use of the raw data and correction for the sound velocity. The \(G_{cal}\) value can then be established together with an estimation of the absorption.
Before introducing the sphere, the echogram noise level should be measured to make sure that there is sufficient SNR for the sphere measurement; clear water should be used. Note also that the measured \(A\) is S+N (Signal + Noise). At low SNR, the noise must be subtracted in the linear domain in order to find the signal strength, S, of the target. This also applies to volume scattering estimates so field data must be corrected in this way.
Updated