Modellizzazione in Mar Ionio

Transcription

1 Modellizzazione in Mar Ionio Rosario Grammauta 1, Salvatore Viola 2, (1) IAMC-CNR UO Granitola, Campobello di Mazara (TP), Italy, (2) INFN - Laboratori Nazionali del Sud, Catania,,Italy rosario.grammauta@iamc.cnr.it 1

2 Summary Spatial distribution of the acoustic noise levels caused by shipping traffic in two areas of the strait of Sicily. Area of Capo Granitola 2

3 LAT [deg] Summary The first study area, the Gulf of Catania, is affected by intense shipping traffic due to the presence of three important harbors: touristic, commercial and military. Calculation of the vessels position on a grid having mesh size 2000m x 2000m, with a time resolution of 30 minutes Study area Min Longitude: 15.0 Max Longitude: 15.9 Min Latitude: 36.8 Max Latitude: 37.8 Analyzed AIS data Time period: November 2012 November 2013 Number of vessels: 3621 Total hour vessels: Gulf of Catania Minute LON [deg] Cumulative minutes of ship traffic calculated in a 2000m x 2000m grid from AIS data 0

4 LAT [deg] Summary Area of Capo Granitola The second area, instead, is the Area of Capo Granitola where the massive maritime traffic is attributable to the Strait of Sicily. Calculation of the vessels position on a grid having mesh size 2000m x 2000m, with a time resolution of 30 minutes Study area Min Longitude: 12.0 Max Longitude: 13.6 Min Latitude: 36.6 Max Latitude: 37.8 Analyzed AIS data Time period: November 2012 November 2013 Number of vessels: Total hour vessels: Area of Capo Granitola Minute LON [deg] Cumulative minutes of ship traffic calculated in a 2000m x 2000m grid from AIS data 0

5 Modelling underwater noise from AIS data The acoustic noise induced by ship traffic in the Gulf of Catania has been estimated with a fast custom-written MATLAB script from the data collected by the AIS receiver installed at the INFN-LNS. Script steps: 1. Calculation of the vessels position on a grid having mesh size 100m x 100m, with a time resolution of 10 minutes. 2. Estimation of the source spectral density for each vessel according to the Research Ambient Noise Directionality model RANDI 3.1 (depending on speed, length of the vessel). 3. Evaluation of the effects of the surface-dipole interference on radiated sound power. 4. Application of an easy geometric spreading model accounting spherical spreading to the maximum water depth along the modeling radius, and cylindrical spreading for the remainder of the radius. 5. Computation of the frequency-dependent absorption

6 Modelling underwater noise from AIS data Calculation of the vessels position on a grid having mesh size 100m x 100m, with a time resolution of 10 minutes. Study area Min Longitude: 15 Max Longitude: 16 Min Latitude: 36.8 Max Latitude: 37.8 Analyzed AIS data Time period: 1st October, 2012 to 28th February, 2013 The pink triangle indicates the location of the NEMO-SN1 observatory ( N, E, depth 2100 m). Gulf of Capo Catania Cumulative Minute Cumulative minutes of ship traffic calculated in a 100m x 100m grid

7 Modelling underwater noise from AIS data Estimation of the source spectral density for each vessel according to the Research Ambient Noise Directionality model RANDI 3.1 (depending on speed, length of the vessel). In this formula, cv and cl are power-law coefficients for speed and length (taken to be 6 and 2, respectively), v0 is the reference speed (12 kt), l0 is the reference length (300 ft), Ls0(f) is a mean reference spectrum, and g(f,l) is an additional length-- dependent correction to the Ross model (Breeding et al., 1996). reeding, J. E., Pflug, L. A., Bradley, M., Herbert, M., and Wooten, M. (1994). RANDI 3.1 User's Guide: Naval Research Laboratory. reeding, J. E., Pflug, L. A., Bradley, M., Walrod, M. H., and McBride, W. (1996). Research Ambient Noise Directionality (RANDI) 3.1 hysics Description: Planning System Incorporated.

8 Modelling underwater noise from AIS data Evaluation of the effects of the surface-dipole interference on radiated sound power. To account for the effect of surface-dipole interference on radiated sound power, source levels at wavelengths greater than four times the source depth were attenuated according to the relation given by Brekhovskikh and Lysanov

9 Modelling underwater noise from AIS data Application of an easy geometric spreading model accounting spherical spreading to the maximum water depth along the modeling radius, and cylindrical spreading for the remainder of the radius. 9

10 Modelling underwater noise from AIS data Computation of the frequency-dependent absorption TL = αr due to the relaxation of boric acid B(OH3) and magnesium sulphate MgSO4 molecules, and the shear & bulk viscosity of pure water

11 Validation of Model NEMO-SN1 Observatory Long-term background noise measurements in the Gulf of Catania Evaluation of the background acoustic noise from ship traffic Comparison between the AIS-derived noise levels and the real acoustic data

12 The NEMO-SN1 Observatory and the Catania node of EMSO NEMO - SN1 is the first-established node of EMSO (European Multidisciplinary Seafloor Observatory), one of the incoming European large-scale research infrastructure included since 2006 in the Roadmap of the ESFRI (European Strategy Forum on Research Infrastructures). NEMO SN1 INFN LNS shore lab Up: geographic location of the NEMO SN1 multidisciplinary observatory Right: NEMO SN1 operative underwater Low frequency hydrophone NEMO-SN1 seafloor observatory has been operating from June, 2012 to May, 2013 in the Western Ionian Sea (East Sicily, N E), at about 2100 m water depth. 3

13 NEMO-SN1 low-frequency hydrophone Underwater acoustic noise in the Gulf of Catania was measured through a seismic hydrophone, model SMID DT- 405D (10 Hz<f<1 KHz). Acoustic data were collected from June 2012 to May 2013, 24 hours per day, at a sampling frequency of 2 khz, saved in 10 min long files and stored at the INFN LNS. PREAMP DIGITIZER DATA OUT DATA IN (CONTROL) PPS GPS (NMEA) Ch0=+60 db Ch1=+30 db High Gain Channel (+60 db) About min files stored Low Gain Channel (+30 db) 4

14 Underwater background noise in the Gulf of Catania Distribution of the mean values of the PSD within each 10 min long recording (binning: 1dB re 1 µpa 2 /Hz) for the whole NEMO-SN1 acoustic dataset (46040 recordings). PSD of each file computed using Welch's overlapped segment averaging estimator (window: 2048 samples, overlap 50%). The gure also includes the curves of the 50th, 90th and 10th percentiles of the mean values of the PSD calculated on the whole acoustic dataset. June May recordings

15 AIS data acquisition at the INFN-LNS laboratory The Automatic Identification System (AIS) is an automatic tracking system used by ships for identify and locate vessels. It gives information such as position, mmsi, speed and length. INFN LNS CATANIA AIS data displayed in real time on a Google map web page INFN-LNS STORAGE 6

16 Comparison between AIS-derived and real acoustic data 1/3 SPL within the 31.5 Hz, 63 Hz, 125 Hz and 250 Hz octave bands at the NEMO-SN1 location (red lines) was simulated with a time resolution of 10 minutes and it was compared with the SPL measured from NEMO-SN1 recordings (blue lines). A minimal background noise given by the smallest values of SPL obtained by the NEMO-SN1 data was considered in the simulation. NEMO-SN1 recordings Simulation 12

17 Comparison between AIS-derived and real acoustic data 2/3 Distribution of the difference (in db) between the estimated SPL (db re 1 µpa) at the NEMO-SN1 location and the SPL measured from NEMO-SN1 recordings (calculated with a time resolution of 10 min) for the 63 Hz and 125 Hz centered octave bands. Period between 1 st October, 2012 and the 28 th February, 2013.

18 Comparison between AIS-derived and real acoustic data 3/3

19 Shipping noise evaluation at different depths Noise levels have been estimated at different frequencies and depths on a grid having mesh size 100m x 100m, with a time resolution of 10 minutes. Mean PSD at 100 Hz Mean PSD at 100 Hz Mean PSD at 100 Hz NEMO-SN1 NEMO-SN1 NEMO-SN1 Estimated mean acoustic PSD at 100 Hz (db re 1 µpa 2 /Hz) induced by the ship traffic in the study area for December 2012 at three different depths

20 Results Estimated mean acoustic PSD at 63 Hz (db re 1 µpa 2 /Hz) induced by the ship traffic in the study area for all 2013 at 5m of depth. Area of Capo Granitola Area of Capo Granitola Gulf of Catania 15

21 Results Estimated mean acoustic PSD at 125 Hz (db re 1 µpa 2 /Hz) induced by the ship traffic in the study area for all 2013 at 5m of depth. Area of Capo Granitola Area of Capo Granitola Gulf of Catania 15

22 db re 1 µpa 2 /Hz Results Comparison: Estimated, with Model, mean acoustic PSD at 63 Hz in the two areas of studio 115 Area of Capo Granitola db 100 Gulf of Catania Winter Spring Summer Autumn Titolo asse Granitola 63 Hz Catania 63 Hz Mean 2013 Mean 2013

23 source spectral density for each vessel Application of Model we use the model to predict what would happen if the ships would change its speed the intensity of the source is dependent on the speed of the ship decreasing the speed of the ships also decreases the intensity of the signal emitted by ships

24 db re 1 µpa2/hz Comparison acoustic PSD Granitola after changed speed ship Area of Capo Granitola speed vessels reduced by 10 % Area of Capo Granitola 113 Comparison mean acoustic PSD at 63 Hz Granitola (Year 2013) db Winter Spring Summer Autumn Granitola 63 Hz 2013 Granitola 63 Hz reduced Mean 2013 reduced

25 db re 1 µpa 2 /Hz Comparison acoustic PSD Catania after changed speed ship Gulf of Catania speed vessels reduced by 10 % Gulf of Catania 99 Comparison mean acoustic PSD Catania (Year 2013) db Winter Spring Summer Autumn Catania 63 Hz Mean 2013 Catania 63 Hz reduced Mean 2013 reduced

26 db re 1 µpa 2 /Hz 115 Management indication Estimated mean acoustic PSD at 63 Hz in the two areas for Area of Capo Granitola 105 Marine Strategy 12 db Gulf of Catania 90 Winter Spring Summer Autumn Granitola 63 Hz Catania 63 Hz Mean 2013 Mean 2013 marine strategy 1

27 db re 1 µpa2/hz Management indication Hypothesized mean acoustic PSD at 63 Hz in the two areas for the speed of the ships is decreased by 10 % Area of Capo Granitola Marine Strategy 8 db Gulf of Catania Winter Spring Summer Autumn marine strategy Granitola 63 Hz Mean 2013 Catania 63 Hz Mean 2013

28 Management indication Area of Capo Granitola If speed vessels reduced by 10 % Area of Capo Granitola If speed vessels reduced by 20 % Area of Capo Granitola (db re 1 µpa 2 /Hz) Area in the threshold of the Marine Strategy

29 CONCLUSION Conclusion and Future The model is a useful tool to evaluate the spatial distributions of induced noise from vessel traffic. It is able to generate real-time maps of noise using exclusively of AIS data; The model can be a useful tool for marine traffic control and for establish new mitigation rules in order to reduce the marine noise (Europea Marine Strategy). FUTURE Improve the model in shallow water Improve the model by adding information about the tonnage and draft of the ships; Improve the validation of the model increasing the data collection point (Artic, Antartic, etc).

30 NOISE MAP PROJECT Thanks for your attention

31 Suitability of five commonly used propagation models for different water depths and frequencies 1