I use this in the context of James Hansen 'too much CO2 in the atmosphere and Earth becomes like Venus' In Nature Geoscience, this (I just link to their page): http://www.nature.com/ngeo/journal/v6/n8/full/ngeo1892.html First the bad news: solar brightening will do this to the Earth in 500 to 1000 million years. The previous models had that much further into the future. But the good news: according to Goldblatt et al. models here, prompt fossil-fuel burning can probably not 'Venusify' the Earth. They suggest 30,000 ppm would be required. Thirty parts per thousand. Three percent CO2! The stuff would put us all to sleep before we could burn up to that level, even if enough 'fossil' can be scrounged. So all we have to worry about is a few degrees warmer, rain falling somewhere else instead, a new (lower pH) ocean ecosystem, and a bit less continental area around the edges. Carry on, then. It was beyond the scope of Goldblatt et al. to consider whether a few degrees warmer could destabilize the methane hydrates in the ocean. That could cause 'the big change', perhaps, but we really have no idea how much methane is down there. +++ The way to get this paper free is 1. follow the link 2. click corresponding author 3. provide your name, email address and check on 'send me a copy' 4. adding a polite message in the text box is optional 5. be impolite, you probably won't get the $32 thing for free.
There are not many scientists who conclude as Hansen does, that a runaway greenhouse could be achieved by plausible fossil-C burn rates. So in that sense, the model here is not new. They do put a number on it, which is new (I think). Their current model is one-dimensional, which they plan to upgrade in the future. I do think that exoplanet research could make fuller use of the current crop of climate models. One standing objection to climate modeling is that we have an unreplicated system (Earth). With adequate spectroscopy on adequate space-borne telescopes, the chemistry of exoplanet atmospheres could be observed. Bam, there's your replication. I suppose we need one or two orders of magnitude improvement in the telescope technology to pull this off. Back in the days before Moore's law, that would have been seen as a big problem. Not saying that space telescopes will ever be low cost, however.
