Scientists like Doron Nof, Ian McKeague, and Nathan Paldorcan say it much better than I can.
"Lake Kinneret (the Sea of Galilee) is a small freshwater lake (148 km2 and a mean depth of 20 m) situated in northern Israel. Throughout recent history there have been no known records of a total ice formation on its top. Furthermore, given that convection requires an initial cooling of the entire lake down to 4 °C, it is difficult to imagine how such a low-latitude lake, presently subject to two-digit temperatures during the winter, could ever freeze. Lake Kinneret is, however, unique in the sense that there are dense (warm and salty) springs along its western shore. The dynamics of the regions adjacent to these springs are investigated using a one-dimensional nonlinear analytical ice model, a paleoceanographic record of the sea surface temperature of the Mediterranean Sea, and a statistical model. We show that, because the water directly above the plume created by the salty springs does not convect when it is cooled down to 4 °C, freezing of the region directly above the salty springs was possible during periods when the climate in the region was somewhat cooler than it is today. We refer to this localized freezing situation as ‘springs ice’.
The analytical ice-model involves a slowly varying approach where the ice is part of a thin fresh and cold layer floating on top of the salty and warm spring water below. During the ice formation process, the ice is cooled by the atmosphere above and warmed by the spring water below. The plumes created by the springs have a length scale of 30 m, and it is argued that, during the Younger Dryas when the air temperature in the region was probably 7 °C or more cooler than today, ‘springs ice’ (thick enough to support human weight) was formed once every 27 years or less. During the cold events 1500 and 2500 years ago (when the atmospheric temperature was 3 °C or more lower than today) springs ice occurred about once in 160 years or less. Since the duration of these cold events is of the same order as the springs ice recurrence time, there is a substantial chance that at least one springs ice occurred during these cooler periods. With today's climate, the likelihood of a springs ice is virtually zero (i.e., once in more than 10,000 years).
One set of those springs associated with the freezing is situated in Tabgha, an area where many archeological features associated with Jesus Christ have been found. On this basis, it is proposed that the unusual local freezing process might have provided an origin to the story that Christ walked on water. Since the springs ice is relatively small, a person standing or walking on it may appear to an observer situated some distance away to be ‘walking on water’. This is particularly true if it rained after the ice was formed (because rain smoothes out the ice’s surface). Whether this happened or not is an issue for religion scholars, archeologists, anthropologists, and believers to decide on. As natural scientists, we merely point out that unique freezing processes probably happened in that region several times during the last 12,000 years." Click here for the website I pulled this from.
