We present the Environmental Scenario Search Engine (ESSE), a set of algorithms and software tools for distributed querying and mining large environmental data archives. The principal requirement of the ESSE system is to allow the user to query the data in meaningful “human linguistic” terms. The mapping between human language and computer systems involves fuzzy logic. We use a data resource web service abstraction layer to virtualize spatio-temporal databases, providing our search engine with time-series of environmental parameters. The data resource interface is implemented as a set of OGSA-DAI components with simple input and output XML schemes. Time-series selected from a data resource in XML format can be mined for environmental events by the ESSE or used after XSLT transformation by other clients, Microsoft Excel 2003 being one of the examples.