Leaf maintenance respiration becomes negative at high temperatures under certain parameter combinations

[JessicaNeedham]

Under certain parameter combinations and high temperature extremes, maintenance respiration can become negative using the Atkin et al. 2017 respiration scheme - see notebook here. This happens because the r2 parameter is negative, so if temperature gets hot enough, the whole term can become negative:

fates/biogeophys/FatesPlantRespPhotosynthMod.F90

Line 2181 in 73f31ab

r_t_ref = nscaler * (r_0 + r_1 * lnc_top + r_2 * max(0._r8, (tgrowth - tfrz) ))

This is only going to occur at very high temperatures (above ~30 C) and with low values of r0. But given r0 is a parameter that users might adjust, and we are running future climate scenarios, it seems like we should either put some cap on leaf maintenance respiration or crash the model. I’m in favour of wrapping the whole thing in a max(0, lmr) term with a warning that lmr has gone to zero, but curious what others think. We could also have some warnings when parameters are read in, that a low value of r0 might cause lmr to go to zero/negative at high temperatures.