BIO-OPTICAL DATA INTEGRATION BASED ON A 4D DATABASE SYSTEM APPROACH

Abstract

Bio-optical characterization of water bodies requires spatio-temporal data about Inherent Optical Properties and Apparent Optical Properties which allow the comprehension of underwater light field aiming at the development of models for monitoring water quality. Measurements are taken to represent optical properties along a column of water, and then the spectral data must be related to depth. However, the spatial positions of measurement may differ since collecting instruments vary. In addition, the records should not refer to the same wavelengths. Additional difficulty is that distinct instruments store data in different formats. A data integration approach is needed to make these large and multi source data sets suitable for analysis. Thus, it becomes possible, even automatically, semi-empirical models evaluation, preceded by preliminary tasks of quality control. In this work it is presented a solution, in the stated scenario, based on spatial - geographic - database approach with the adoption of an object relational Database Management System - DBMS - due to the possibilities to represent all data collected in the field, in conjunction with data obtained by laboratory analysis and Remote Sensing images that have been taken at the time of field data collection. This data integration approach leads to a 4D representation since that its coordinate system includes 3D spatial coordinates - planimetric and depth - and the time when each data was taken. It was adopted PostgreSQL DBMS extended by PostGIS module to provide abilities to manage spatial/geospatial data. It was developed a prototype which has the mainly tools an analyst needs to prepare the data sets for analysis.