INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT CROATIA since

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 INSTITUTE OF OCEANOGRAPHY AND FISHERIES Setaliste Ivana Mestrovica 63, 21000 Split; Damjana Jude 12, 20000 Dubrovnik MEDAS SYSTEM FOR ARCHIVING, VISUALISATION AND VALIDATION OF OCEANOGRAPHIC DATA Dadic, V. and D. Ivankovic Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 1

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Database users Croatian National Monitoring Programme Web browser Use Systematic Research of the Adriatic Sea as a Base for Sustainable Development of the Republic of Croatia Application server In Frame of MEDAS Croatian Institutions Database with web interface Marine Environmental Database of the Adriatic Sea (MEDAS) Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 2

3 INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Database structure MARINE PHYSICS • SEA LEVEL • SEA CURRENTS • CTD • METEOROLOGY META DATA • ROSCOP • INSTRUMENS • RESPONSIBLE PERSONS • CLASSICAL OCEANOGRAPHY • (T, S, Ph, o 2, sea surface meteorology) MEDAS FISHERIES • COASTAL FISHERIES • PELAGIC FISHERIES • DEMERSAL FISHERIES • EGGS, LARVES, JUVENILS • CATCH STATISTICS • AQUACULTURE INTERFACE • GIS LAYERS • EXCHANGE FORMATS MARINE CHEMISTRY • NUTRIENTS • HEAVY METALS • P A T • P A H DATABASE BASIC LAYERS • COASTAL LINE • BATHYMETRY • WRECKS … MARINE BIOLOGY • ZOOPLANCTON • ZOOBENTHOS • PHYTOPLANCTON • MICROBIOLOGY Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

4 INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Database capability Online stations (real time) Mapping tool Cruise summary reports MEDAS (near real time – GPRS) Data in digital form Data visualisation database Manually inserted data Data for running models Search criteria, grouping data, statistics Data export: • MEDAR / MEDATLAS • Excel • Ocean Data View • Surfer • Arc GIS 8. 1 Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 MEDAS database - history Distribution of measuring stations archived in MEDAS database Ø ORACLE 5 + Forms 3 – mapping tool: non Ø ORACLE 7 + Forms 5 - mapping tool : C++ Ø Oracle 9 i + Oracle Application server - Mapping tool : Java applet Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 5

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Why this technologies? ORACLE 9 i: Ø Stability, security Ø Lack of stuff (few men's show) ORACLE Application Server: Ø No need for client side software (only browser) Ø Accessibility (different institutions, ship – GPRS) Ø Easy web publishing (default) Java applet – mapping tool and data visualisation: Ø Portability (cross platform) Ø Use of Client side resources (no need for powerful server ) Disadvantages: Ø Software license cost Ø Instability and bugs of some Java versions Ø Sometimes need for manually JVM install Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 6

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 MEDAS database web interface Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 7

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Importance of good search criteria Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 8

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Input forms ØData input trough web forms (Croatian language only) ØAuthorisation requested ØBasic input validation ØSecure protocol (https) Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 9

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Cruise report example Ø Cruise information with mapping tool Ø Frames organised Ø Form – applet interaction Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 10

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Importance of Where, What and Who Where & When Who What Ocean Biodiversity Informatics Expended GF 3 based organisation International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 11

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Real-time circulation model Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 12

Real – time data (automatic stations) INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Station PUNTA JURANA Øin front of the main building of the Institute Ø 12 meteorological oceanographic parameters (nine air and three sea parameters) Station VELI RAT Ømeasurements every 10 minutes 24 hours a day ØMeteorological sensors on the top of lighthouse Ø 11 meteorological - oceanographic parameters (nine air and two sea parameters) ØMeasurements every 10 minutes 24 hours a day, data refresh interval – 1 hour Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 13

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 About data Data sources in MEDAS database Øperiod almost a century (1907– 2004) Øcountries: Croatia, Austria, ex Yugoslavia, Italy, USA, Russia, France, Germany and Slovenia Oceanographic data (characteristics): Ømeasured by research and other vessels Ørandomly distributed stations (not regular in time and space) Ømeasured at various sea levels Ødifferent methods and instruments (different quality of data) Bathimetry and coastline of the Adriatic Sea Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 14

Quality Control and Data Processing INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Data quality control in MEDAS database includes: ØDuplicate data elimination ØRange checking of data ØStatistical checking of data ØStability checking of data. Data profiles (sets) Ø 51769 temperature Ø 32562 salinity profiles Ø 5840 oxygen Ø 2925 p. H Ø 13835 nutrients Ø 3345 chlorophill_a Procedure: two steps QC 1 first step QC 2 second step Ø 1753 plankton (zoo and phyto) Ø 1218 fish trawler data Ø 785 echosounding samples Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 15

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Cases when data excluded from processing CAST CRUISE 0 - No QC 1 - Correct value 2 - Suspect value 3 - Out of climatological range 4 - Interpolated value 5 - wrong value 9 - missing value UNKNOWN DATE DATABASE WRONG POSITION OF STATION EXCLUDE WHOLE DATA CAST NO DEPTH EXCLUDE A DATA FROM CAST Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 16

Classical Oceanographic Data Inventory INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 The most complete data sets: 1907 -2004 Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 17

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 BOT Number of measuring data at different levels CTD MBT XBT Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 18

451 stations with wrong position Split 327 resolved 124 unresolved Makarsk Brač a Hvar Ploče INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 An example of spatial distribution of received oceanographic stations in wide area City of Split Korčula Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 19

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Example: Result of converting from pressure to depth and vice-versa Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 20

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Number of data of temperature at 5 different levels by season Season a) a) Original data b) Data in interval of 3 sigma of average c) Data inside 1 sigma of average Season b) Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 c) 21

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Seasonal presentation of total amount of temperature data Fall Winter Summer Spring Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 22

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Sea surface temperature sorted by time from 1900 up to 2000 Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 23

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Definition of areas in the Adriatic Sea for calculation climtologcal ranges of classical oceanographic parameters (I) (II) (IV) Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 28 one degree squares and 4 sub-regions in the Adriatic Sea 24

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Data processing is done in the next steps in aim to get output data in suitable form for analysis Calculation of basic statistics (number of samples, mean value, standard deviation, min and max values) Procedure: Interpolation of measuring data on 41 standard oceanographic levels using Newton method of finite differences modified by reference RR curves Semiautomatic Visualisation Calculation of basic statistics on standard levels Interpolation by kriging method and graphical presentation by mapping tool No OK YES Finish Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 25

Standard depth Adriatic Sea: 41 standard levels Max distance between two outer levels Max distance between two inner levels Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 Number of standard tween levels Depth (m) INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 26

27 Depth (m) INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Newton interpolation of finite differences (nth order) Newton interpolation of finite differences (2 th order) through 3 points modified by R-R reference curve Comparison of the three different methods of interpolation data on standard levels Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 Temperature (o. C)

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Testing of parameters in reference curves: m=1 -22 n=1, 10 Optimal values: m=1. 6, n=0. 6 Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 28

29 INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Method of analysis of spatially distributed data Kriging method has been used as optimal method of interpolation of randomly distributed data in space because it provides the best linear unbiased estimate (BLUE) of the variable at a given point. It is an exact interpolator in the sense that interpolated values, or best local average, will coincidence with the values at the data points. The kriging method is valid if there is continuity among dana, which is testing by variogram. Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, Semivariance CROATIA, since 1930 Investigation of spatial continuity of temperature by variance Semivariance Winter Spring Summer Autumn Number of pairs Continuity Discontinuity Winter Spring Summer Autumn Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 Distance between pair of stations (*60 Nm) 30

31 INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Quality controlled data All stations (1900 -2004) Spatial distribution of oxygen stations Oxigen stations 37439 drops 7845 drops Spatial distribution of oceanographic measuring stations Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Filed of spatially distribute sea surface temperature derived from original data and data partly passed QC procedure for the period 1907 -2003 Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 32

33 INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Averaged values and standard deviation of salinity in 1 deg. squares (0 -5 meters) for the period 1907 -2004 Square includes: - coastal area - open sea (need to use smaller size square) Possible choosing different size of squares Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Result of data analysis of T, S, O 2, p. H and nutrient parameters 34 It was concluded: Ø 41 standard levels in the Adriatic Sea was defined as enough for qualitative climatological analysis of T, S, O 2, p. H and nutrients in the Adriatic Sea Ø climatological range of oceanographic parameters in the Adriatic sea for above mentioned parameters was defined Ø spatial continuity of oceanographic parameter are from 5 to 7. 5 km depends of season and depth of standard oceanographic level Ø detail structure of spatial distribution for climatological properties of oceanographic parameters has been calculated Ø different layers of the water mass in the Adriatic Sea has been analysed. Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Water mass distribution in the period of low liw inflow (derived from oceanographic data) Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 35

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Water mass distribution in the period of liw inflow intensification (derived from oceanographic data) Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 36

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Heavy metals in dry weight of oyster tissue analysis Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 37

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 38 MEDA S Arc. View USA NODC Tax code Mean concentration of chlorofill_a Ocean Biodiversity Informatics analysis International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Example: spatial analysis of distribution commercialy demersal fish population in the Adriatic Sea Calculation of spatially distribution of relative abundance (kg/km 2) of: HVAR Ø total stock of demersal fish population Ø stock of main commercial spicies from two extensive trawl-surveys carried out in two distinct periods shifted 50 years in time: • Expedition Hvar (1948 -1949) • EU project MEDITS (1996 -1997) MEDA S SURFE R FAO tax code 39 MEDITS Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Spatial distribution of total fish stock (kg/km 2) during HVAR and MEDITS expedition HVAR MEDI TS Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 40

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Spatial distribution of main commercial fishes (kg/km 2) during HVAR and MEDITS expedition Ø Composition changes of demersal stocks assemblage within two surveyed periods should be foreseen as a consequence, either of fishing pressure or environmental changes in the sea. HVAR Ø Decrease of quantity of total as well as the most important commercial fishes of the demersal fish stock Ø The changes in fish abundance are more significant in shallow coastal area then in deeper sea areas MEDIT S Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 41

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Example: planning of assignment of marine areas in the Split-dalmatia County Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 42

INSTITUTE OF OCEANOGRAPHY AND FISHERIES SPLIT, CROATIA, since 1930 Future Improvements: ØTranscode all database thematic parts to web environment ØInput forms (ergonomic) ØWeb visualisation tools ØData quality procedures and tools for biological parameters ØExtend taxonomic codes ØAdd multimedia support ØAdd GIS support Ocean Biodiversity Informatics International Conference on Marine Biodiversity Data Management Hamburg, Germany: 29 November to 1 December 2004 43