Global Climate Monitor

Global Climate Monitor is supported in the design of a data model and a tool to geo-visualize global climate data and climate-environmental indicators. The latter are derived calculations or statistics that are easy to understand and capable of explaining weather patterns on a global scale to any potential user, within or outside the scientific community. It therefore falls in the area of ​​Open Knowledge, as its primary objective is to make complex data easily available.

The data currently available in the geo- viewer comes from multiple sources. TS3.21 version of the CRU database produced by the Climate Research Unit (University of East Anglia) provides precipitation and temperature data from January 1901 to December 2012. It offers data on a monthly basis and with a spatial resolution of half a degree in latitude and longitude. From january 2013 until present, data sources are the Global Precipitation Climatology Centre (GPCC) for precipitation and the Global Historical Climatology Network-Monthly (GHCN-M) version 3.2.1 for global mean temperature.

At the same time, a set of derivative indicators from that database are displayed, offered at monthly, seasonal and annual scale:

Variables & indicators that are presented in the first version of the GCM

*The names in capital are used for the original variables that were served by the Climate Research Unit. Source: Own elaboration.

System architecture

All the applications used to create Global Climate Monitor are open source. The system is based on a multilevel architecture in three layers as follows:

1. Data layer: This layer is the warehouse of the data used by the system. The original database is served in NetCDF by the CRU, which is extracted to text files using the software R. The data server system has a main core component comprising a PostgreSQL relational database server (version 9.2) and its spatial extension PostGIS (version 2.0) for point geometry management. PostgreSQL handles tables with over 90 million records, which shows its great capacity. At a second level, the system has a file server working under the Linux operating system that controls the plain-text files deriving from the netCDF climate data extraction.

2. Business layer: This level functions from the data warehouse to the customer and is supported by two main components: a map server and a web server. GeoServer (version 2.5) is selected to handle the transactions between the database server and the clients through WMS interoperable services. GeoServer v.2.5 obtains data for specified periods and serve them as interoperable services. The points that configure the 0.5 × 0.5 grid in the dataset are displayed with proportionally sized symbols and using also colour ramps. This semiotics provides global coverage so that is really easy to extract a meaning and discover weather patterns.

WMS service can be accessed through this URL.

In addition, NGINX (version 1.1.19) is the web server that integrates the standards for the web content definition. The latter was chosen because it is the best in terms of the number of requests per second that it can serve and because of its memory usage.

3. Client (geo-viewer): The user layer is the visible component of the system. This level is the viewer itself, which is accessible at the following URL: http://www.globalclimatemonitor.org. This viewer was designed and adapted for our project using the JavaScript code libraries that were developed for the OpenLayer project (version 2.13.1) as an effective scripting language for online mapping.

The display consists of a single html page with a main map area and a collapsible menu on the left: the layer switcher. The layer switcher shows the variables and indicators organized into four categories: monthly, annual, normals ​​and trends. When the user selects a variable and a temporal scale or period, the map and the legend are automatically updated.

The collapsible legend is at the bottom left of the main map area.

On the top right, users can find basic tools to zoom and to move over the map, as well as two information buttons: the top button is used to obtain information on the value of the variable at the point clicked and the other switches the map to show the number of stations within correlation decay distance of grid box for interpolation of the variable currently displayed on the screen. This value can be seen as an indicator of the uncertainty of the interpolation.

The user can choose between different base layers using another layer selector placed on top of the map. On its left, there is a tab that displays the geographic coordinates of the mouse location, as it is moved over the map.

On the top left there are three buttons used to open the animation of normals window, the download menu and the charts menu.

General information about the data source, license and contact is shown at the bottom of the map area.