As the number of sensors and quantity of collected data increases for field measurements, there has been much interest in sensor networks. The scientific community faces tremendous challenges in the collection, dissemination, and archiving of vast amounts of raw and processed data. In large field experiments such as high-resolution urban dispersion studies or global climate investigations, data can originate from various types of sensors, sources, and computing processes. Also, data collections can be sporadic and their publication can materialize in an assortment of formats and through dissimilar channels. To resolve many of these challenges, we propose a sophisticated yet simple and elegant architecture called a SensorSpaces ^TM system based on the concept of “tuplespaces” (Dr. David Gelernter, Yale University). A SensorSpaces system provides a virtual space to collect and disseminate sensor data objects that are well defined and easily shared. This concept provides loose coupling of sensors, systems, and models in a networked environment. The objective is to demonstrate that tuplespaces implementations can greatly simplify acquisition and distribution of field measurements. A model of a SensorSpaces system to collect and archive field measurements will be presented. Issues regarding sensor interfaces, legacy instrumentation, wireless communications, data base management, scalability of the number of sensors, distributed processing of loosely coupled systems, and performance of the data acquisition system will be discussed.