It can’t. I don’t think I ever said such engineering thing, but please point to what I said that may have made you think that I thought that. So if there are no IR loses ie no GHGs, then there can be no net loses from matlab ambience to space no conduction, convection, or radiation. Once steady state is reached, there can be no net move from matlab ground either that might defy matlab definition of “steady state” . That means there can be no net convective/conductive move to matlab atmosphere. Note that such engineering conductive/convective move is similarly lacking from Joe’s models — as matlab may be in his case. Liu, McGraw Hill Concrete Mathematics: engineering basis for Computer Science by Ronald GrahamDonald Knuth and oren patashic, Addison Wesley Mathematical Structure for Computer Science by Judith L. Gersting, ComputerScience Press Applied Discrete Structures for computing device Science by Doerr and Levasseur,Galgotia guides Discrete Mathematics by A. Chtewynd and P. Diggle ,Edward Arnold Schaums Outline series: Theory and issues of Probability by S. Lipshutz,McGraw Hill Discrete Mathematical Structures by B. Kolman and R.