Use of TimescaleDB as a database for ocean-meteorological data storage

Abstract

The monitoring of the environment is essential to know and study the diferent phenomena that occur in the natural environment. [1]. Atmospheric and oceanographic data sets share many characteristics. They can be very large; many cover limited periods of time and have a limited spatial extent; gaps (lack of data) and outliers are common; The spatial distribution of several observation networks is uneven; and, often, the time series of data are not homogeneous. The data sets contain variables that, in general, are not independent in time or space; therefore, most of the variables must be seen within a multivariate context [2]. A time series database [3] (TSDB) is a database optimized for timestamp data. These databases have the ability to provide queries with subsamples, gap flling or aggregations throughout the time series, must be stored efciently to be inserted and retrieved quickly. TimescaleDB [4] is implemented as an extension on PostgreSQL, and exposes what look like singular tables, called hypertables, that are actually an abstraction or a virtual view of many individual tables holding the data, called chunks (created by partitioning the hypertable’s data into one or multiple dimensions). We will proceed to perform a series of tests to test the performance of TimeScale and compare it with other databases of the same style. Small comparisons will be made both in writing and reading data. Special emphasis will be placed on reading large datasets.