Tom talked about the various sea surface temperature monitoring datasets in an earlier post. (The images will be restored soon--we're still upgrading old posts on our new climate.gov website.) They all have different inputs (the actual measurements), methods of going from single-point measurements to a regular grid, and even reporting periods--daily, weekly, or monthly. There are usually differences between the temperatures in the various datasets, and a difference of 0.2 degrees isn't unusual. We do pay close attention to the temperatures, though, and sometimes it becomes evident that one of them has an error. We think this is just the "normal" uncertainty in the data, but we'll keep an eye on it.