This paper examines the use of Elasticsearch for data warehousing and analyses of geo-referenced sensor data. Elasticsearch has several advantages compared to its direct competitors. For example, it is capable of handling time series, spatial data, and objects. Moreover, it is natively connected with the data shippers Beats, Logstash, and the visualisation tool Kibana. This paper proposes a method to implement and query multidimensional models in Elasticsearch. No prior work has evaluated Elasticsearch for data warehouses and analytical queries, especially for sensor environmental data. This paper therefore also presents extensive experiments to evaluate its querying performance. The proposed approach is applied to the analysis of sensor data used in the context of CEBA, an environmental cloud solution developed to collect, store, and analyse environmental data. An experimental performance analysis is also provided.