Methods of Oil Spills Hydrophysical Monitoring. Методы гидрофизического мониторинга нефтеразливов

Transcription

1 Valery Vitko, Arctic Public Academy of Science Russian State Hydrometeorological University В. Б. Митько, Арктическая общественная академия наук, Российский государственный гидрометеорологический университет Methods of Oil Spills Hydrophysical Monitoring Методы гидрофизического мониторинга нефтеразливов Saint-Petersburg Санкт-Петербург 2009

2 Here are presented some results and recommendations of the research, fulfilled in frame of project Actual Problems of Socium Safety Russian Academy of Natural Sciences (Section of Geo-policy and Safety) on base of theoretical and experimental works of: Arctic Public Academy of Sciences; Agency of Science-Intensive and Innovative Technologies Prognoz-Nord ; Research Institute of Radio-electronic System of Extraordinary Situation Forecasting Prognoz of Saint-Petersburg State Electro-technical technical University; Laboratory of Remote Monitoring of Russian State Hydrometeorological University.

3 The objective of the projects The maintenance of local administrations with the inexpensive tool of monitoring of the environment and emergency situations for the control and prediction of the main parameters describing the situation of the environment in Integrated Coastal Area Management (ICAM), connected with oil spills generation

4 STRUCTURE OF NEVA BIGHT ECOLOGICAL MONITORING SYSTEM

5 The Disposition of Stations in Experiment of Prognoz Prognoz ETU

6 Map of Measurement

7 Solving Methods For initial Data Processing Algorithms are developed which include Prolonged Signals Collection for Contrast Radar Images Obtaining, finding the logarithm of received Signal Amplitudes, Excision of invariable Image Elements and Digital Conversion. Digital Data Processing for contrast Oil Spill Images obtaining, their Classification, Geometric Proportions Determination and Mapping with developed Software

8 Signal s Spectrum

9 Signal s Spectrum

10 Range Rate

11 Range Rate

12 Map of RSHU Experiment Area

13 Radar Image after initial Processing Scale: 1.5 mile with 72 Images Imposition for Slicks Stability Illustration

14 Radar Image Processing For Oil Spills Detection 100 m. x 100 m. Resolution is sufficient. Radar Image Analysis starts from Dark Object Boundaries Detection. Its Surface is analyzed and a Histogram is constructed in which X axis indicates Gray Intensity Level and Y axis indicates amount of Pixels. Typical Form of such a Histogram is presented in following Figure

15 Histograms а) Histogram of Oil Spill b) Histogram of Natural Slick

16 Histogram Processing Oil Spill Histogram has two Peaks. The Smaller is located in the Area of Median Backscattering Value of the Dark Object. The Larger is located in the Area of Background Median Backscattering Value. The local Minimum between the two Peaks is used for Image Fragmentation. For Fragmentation Procedure the darkest Pixel is chosen as a Start Point and then an Area is grown around it until this Area is less than a Threshold calculated from the local Minimum of the Histogram

17 Spill Image Processing Algorithm Oil Spills have lower Complexity and they are thin in Comparison with Wind Slicks. Median Gradient Values of Oil Slicks along Borders are higher than those of Wind Slicks which are usually larger than Oil Spills. For Oil Spills Detection in MATLAB a software was developed. The main Window of the program is presented in Figure. The Program Interface provides for resolution input (100 m. x 100 m.). For Histogram Output an Isolated Button is used. 1) Detection Threshold (Pixel Intensity) is derived from the Histogram for marking out an Area-Spill Candidate.

18 The main Program Window

19 Proposed Structure of Hydroacoustical Experiment in Arctic Regions Covered by Ices

20 Main Correlations Formula of Marsh has view: W=1-0,458(3fH)3/2(3H)1/10 )1/10, Where -ray grazing angle, rad, f-operating frequency,, кhzк Hz, Н-average the wave height. Reflection coefficient depends on grazing angle from 0 to 40 db. Oil on surface changes coefficient of liquid tension and degree of wind waving. It is necessary hydrophisical attestaition of regions with subsequent periodical of reflection coefficient measuring

21 Mapping The created Macros with dkart Navigator Redactor is attached to the given Region of the Electron Map. Reference Points are Coastal Line Images or other constant Objects in the survey field. The Printer prints the Map Paper-Copy.

22 Conclusions Developed Algorithms provide for Solution of Oil Spill and Wind Slicks Marking up Problem and allow to classify Oil Spills and Wind Slicks. In Presence of thicker films the higher Contrast will be observed and Oil Spill marking up Probability will be higher.