Computer models and remote sensing missions often generate datasets that are too large to store and analyze on desktop computers or even on local compute clusters. A traditional approach to exploring such datasets is to run the analysis program on a remote machine and display the output on the desktop computer via X windows. The tunneling option of the secure shell connection can be used to facilitate remote display of graphics generated by analysis software like Matlab, IDL, NCL etc. More recently, purely browser-based visualization tools are becoming more popular for graphical data analysis and visualization. These tools extend beyond desktop and laptop computers to the realm of mobile devices. Javascript toolkits like d3.js support powerful in-browser display options. The IPython Notebook interface combines markup, code, and graphics within a single browser-based document, serving as a replacement for the traditional terminal/editor/graphics window development environment. The notebook supports versatile inline graphics and widgets for remote visualization. This presentation will discuss different alternatives that are currently available for browser-based remote data analysis and discuss their pros and cons in the context of both teaching and research