RT Conference Proceedings
T1 A containerized service for clustering and categorization of weather records in the cloud
A1 Sánchez Gallegos, Dante D.
A1 González Compean, J.L.
A1 Alvarado Barrientos, Susana
A1 Sosa Sosa, Victor Jesus
A1 Vargas Tuxpan, Jose
A1 Carretero Pérez, Jesús
AB This paper presents a containerized service for clustering and categorization of weather records in the cloud. This service considers a scheme of microservices and containers for organizations and end-users to manage/process weather records from the acquisition, passing through the prepossessing and processing stages, to the exhibition of results. In this service, a specialized crawler acquires records that are delivered to a microservice of distributed categorization of weather records, which performs clustering of acquired data (the temperature and precipitation) by spatiotemporal parameters. The clusters found are exhibited in a map by a geoportal where statistic microservice also produce results regression graphs on-the-fly. To evaluate the feasibility of this service, a case study based on 33 years of daily records captured by the Mexican weather station network (EMAS-CONAGUA) has been conducted. Lessons learned in this study about the performance of record acquisition, clustering processing, and mapping exhibition are described in this paper. Examples of utilization of this service revealed that end-users can analyze weather parameters in an efficient, flexible and automatic manner.
PB IEEE
SN 978-1-5386-4152-1
YR 2018
FD 2018-07-11
LK https://hdl.handle.net/10016/33939
UL https://hdl.handle.net/10016/33939
LA eng
NO This work was partially supported by the sectoral fund of research, technological development and innovation in space activities of the Mexican National Council of Science and Technology (CONACYT) and the Mexican Space Agency (AEM), project No.262891.
DS e-Archivo
RD 18 jul. 2024