Ferdinand Colombo, the son of Christopher Columbus, translated his father's ship logs and published them. You can find them in translation in your library under Benjamin Keen, "The Life of the Admiral Christopher Columbus by His Son, Ferdinand. Rutgers Univ. Press. 1959.
In his travels to Jamaica he noted,
"The sky, air, and climate were just the same as in other places; every afternoon there was a rain squall that lasted for about an hour. The Admiral writes that he attributes this to the great forest of that land; he knew from experience that formerly this also occurred in the Canary, Madeira, and Azore Islands, but since the removal of forests that once covered those islands, they do not have so much mist and rain as before."
Well, Chris was being modest about the Iberians felling all the trees on those islands. Iberian settlers also felled all the native citizens of said same islands but unlike the trees, doing away with the indigenous peoples had little to do with the desertification of the islands. The notion that the forests bring rain has persisted, to this day. CED readers want to know: Do the forests bring rain?
The best mechanistic study so far is by Alfred J. Garrett in 1982. A Parameter Study of Interactions Between Convective Clouds, the Convective Boundary Layer, and a Forested Surface. Monthly Weather Review 110:1041-1959. Paper titles are not always exciting. “Forests and Rain” would have attracted a bigger audience.
Garrett concludes his paper with this:
"The simulations also suggest that surface parameters such as soil moisture, forest coverage and transpiration, and surface roughness may affect the formation of convective clouds and rainfall through their effect on boundary-layer growth."
Garrett's goal was to predict air mass thunderstorms and rain. An air mass thunderstorm is one that pops up unrelated to proximity to fronts or to orography. Think, if you will, of the dog days of summer at the Coweeta LTER [90 F daily maxima, 73 F dewpoint temperature]. No, this is not Devil's Island. The Bermuda Highs have "backed in" off the subtropical Atlantic. High pressure in the low latitudes rules the day and the day after and the day after, etc. Frontal systems, which are wimps in summer, except in the grand and glorious year of 1992, are nowhere to be found. Moist maritime tropical air comes ashore. The forests load more water into the air. The forests load hydrocarbons into the air as well. Those hazy, crazy days of summer are here! Pocahontas must surely have known these hazy days as well. In this world of hazy, stable air, eyes scan the late afternoon sky for thunderstorms, rain and relief. If you get one it is, most likely an air mass thunderstorm. These are the systems Garrett wanted to predict.
Garrett used a numerical weather forecasting model and put into it a surface biosphere part to the model. A weather forecasting model with a living surface! Across the Southeast of the U.S. vegetation type was parameterized, a vegetation which could flux mass [H2O] and energy [latent and sensible] into the atmosphere. Forest vegetation coverage ran from 0 to 100% but Garrett only ran his simulations with cover ranging from 40 to 70%. It was a grasslands or forest kind of partitioning of the surface biosphere. Also numerically in his model were stomatal resistance, soil moisture and roughness length.
Garrett found that his weather forecasting model with a biosphere out-performed the forest-free forecasting model used by the National Weather Service in predicting the amount, location and timing of thunderstorms. Thunderstorms over areas with extensive forests occur later in the day (about 3 hours delayed over grass land cover type). Garrett's model is a hypothesis that the kind and condition of surface vegetation makes a difference in the weather realized. He accepts the hypothesis on the basis that his model with vegetation had much higher forecast skill than his model without surface vegetation. That's the way these weather geeks do things. They get in trouble when they try to do manipulative experiments on your and my weather! Without true experiments they call observation programs experiments and predictions hypothesis testing. Better living through better prediction.