An End-to-End System to Enable Quick, Easy and Inexpensive Deployment of Hydrometeorological Stations

Abstract:

The high cost of hydro-meteorological data acquisition, communication and publication systems along with limited qualified human resources is considered as the main reason why hydro-meteorological data collection remains a challenge especially in developing countries. Despite significant advances in sensor network technologies which gave birth to open hardware and software, low-cost (less than $50) and low-power (in the order of a few miliWatts) sensor platforms in the last two decades, sensors and sensor network deployment remains a labor-intensive, time consuming, cumbersome, and thus expensive task. These factors give rise for the need to develop a affordable, simple to deploy, scalable and self-organizing end-to-end (from sensor to publication) system suitable for deployment in such countries.

The design of the envisioned system will consist of a few Sensed-And-Programmed Arduino-based sensor nodes with low-cost sensors measuring parameters relevant to hydrological processes and a Raspberry Pi micro-computer hosting the in-the-field back-end data management. This latter comprises the Python/Django model of the CUAHSI Observations Data Model (ODM) namely DjangODM backed by a PostgreSQL Database Server. We are also developing a Python-based data processing script which will be paired with the data autoloading capability of Django to populate the DjangODM database with the incoming data. To publish the data, the WOFpy (WaterOneFlow Web Services in Python) developed by the Texas Water Development Board for ‘Water Data for Texas’ which can produce WaterML web services from a variety of back-end database installations such as SQLite, MySQL, and PostgreSQL will be used. A step further would be the development of an appealing online visualization tool using Python statistics and analytics tools (Scipy, Numpy, Pandas) showing the spatial distribution of variables across an entire watershed as a time variant layer on top of a basemap.