Abruptmonsoontransitionsasseeninpaleorecords can be explained by moisture-advection feedback Anders Levermanna,b,c,1, Vladimir Petoukhova, Jacob Schewea, and Hans Joachim Schellnhubera,d Abrupt Monsoon Transitions Exist in Paleorecords and Pertinent Models continuity, and zero boundary conditions for vertical velocity at surface and tropopause, yields Paleoclimatic records show abrupt monsoon shifts at various different locations and historic periods (1–5). An important question is whether such transitions are possible in the future (6). To this end, we carved out the physical mechanism for such transitions in a purposefully simple conceptual model (7). Recently, Boos and Storelvmo (8) claimed that introducing adiabatic cooling into our model (7) eliminates these abrupt transitions. This claim is not generally true. As can be seen from their figure 1, Boos and Storelvmo (8) only eliminate abrupt transitions if most of the energy from the rain’s latent heat release is consumed by adiabatic cooling. Although adiabatic cooling exists and is, as shown below, implicitly accounted for in ref. 7, it is not a valid assumption that most of the latent heat release is consumed by this process. By contrast, monsoon circulation is predominantly sustained by latent heat release, as shown in a number of studies (e.g., ref. 9) including figure 2 in ref. 7. Adiabatic Cooling Is Present in Our Model Boos and Storelvmo (8) used a very specific linearized representation of adiabatic cooling based on an approximation of the second horizontal derivative of temperature by a linear function of its first horizontal derivative, dividing velocity scale and length scale (their equations 1−3, S1, and S2). This approximation is crude. Here we show that our model (7) implicitly incorporates adiabatic cooling without eliminating the possibility of abrupt monsoon transitions as have been reported from paleorecords. Our model (7) is based on the entropy equation ds QV , ρ = dt T [1] with s = cp ln ϑ, cp, and ϑ being specific entropy, specific heat, and potential temperature of air, respectively. QV is the net heating rate per volume, and T is kinetic temperature. Vertical integration from Earth’s surface to the tropopause, ztr, and horizontally over the monsoon land region Σ, assuming quasi-stationarity, cp Z !Z X ztr 0 " Z !Z ∇H · ρϑ~ V H dz dσ = X + ztr 0 Z !Z X " ϑ QV ,P dz dσ T ztr 0 " ϑ QV ,R dz dσ, T [2] where H denotes the horizontal plane. QV ,P and QV ,R are the net heating rate per unit volume due to the condensation/precipitation and radiation, respectively. In our simple model, we neglected the sensible heat flux at the land surface. The precipitation heating is nonnegative and the radiation heating is nonpositive throughout the entire troposphere in the continental monsoon regions (10). The deviation of the positive factor ϑ=T from 1 quantifies the contribution of vertical motions to the atmospheric heat balance, i.e., in the adiabatic atmosphere (T = ϑ) vertical motions cannot contribute to the heat balance. Bringing ϑ=T out of the integrals in Eq. 2, one obtains an equation with identical structure to equation 1 of ref. 7, LPSP − « cp W ΔT + RSR = 0, [3] where SP and SR are integral precipitation- and radiationdriven stability parameters. Using the same notation and definitions as in both models (7, 8), one then gets W3 + $ β 2 α # αβ W − RSR = 0, LSP β qO + RSR W − 2 «ρ « c p « ρ cp [4] which has the same general structure as the governing equation 5 in ref. 7 showing the same abrupt transitions. As a consequence, the threshold behavior is not eliminated by adiabatic cooling unless it consumes practically all of the energy of the latent heat release. However, the latter assumption is supported neither by record nor by theory. a Potsdam Institute for Climate Impact Research, 14473 Potsdam, Germany; bLamont-Doherty Earth Observatory, Columbia University, New York, NY 10964; cInstitute of Physics, Potsdam University, 14476 Potsdam, Germany; and dSanta Fe Institute, Santa Fe, NM 87501 Author contributions: A.L. and V.P. designed research; A.L. and V.P. performed research; J.S. and H.J.S. contributed new reagents/analytic tools; and A.L. and V.P. wrote the paper. The authors declare no conflict of interest. 1 To whom correspondence should be addressed. Email: Anders.levermann@pik-potsdam.de. www.pnas.org/cgi/doi/10.1073/pnas.1603130113 PNAS Early Edition | 1 of 2 LETTER LETTER 1 Burns SJ, Fleitmann D, Matter A, Kramers J, Al-Subbary AA (2003) Indian Ocean climate and an absolute chronology over Dansgaard/Oeschger events 9 to 13. Science 301(5638):1365–1367. 2 Wang Y, et al. (2008) Millennial- and orbital-scale changes in the East Asian monsoon over the past 224,000 years. Nature 451(7182):1090–1093. 3 Stager JC, Ryves DB, Chase BM, Pausata FS (2011) Catastrophic drought in the Afro-Asian monsoon region during Heinrich event 1. Science 331(6022): 1299–1302. 4 Gupta AK, Anderson DM, Overpeck JT (2003) Abrupt changes in the Asian southwest monsoon during the Holocene and their links to the North Atlantic Ocean. Nature 421(6921):354–357. 5 Dixit Y, Hodell DA, Petrie CA (2014) Abrupt weakening of the summer monsoon in northwest India ∼4100 yr ago. Geology 42:339–342. 6 Zickfeld K, Knopf B, Petoukhov V, Schellnhuber HJ (2005) Is the Indian summer monsoon stable against global change? Geophys Res Lett 32(15):L15707. 7 Levermann A, Schewe J, Petoukhov V, Held H (2009) Basic mechanism for abrupt monsoon transitions. Proc Natl Acad Sci USA 106(49):20572–20577. 8 Boos WR, Storelvmo T (2016) Near-linear response of mean monsoon strength to a broad range of radiative forcings. Proc Natl Acad Sci USA 113(6):1510–1515. 9 Webster PJ, et al. (1998) Monsoons: Processes, predictability, and the prospects for prediction. J Geophys Res 103(C7):14451−14510. 10 McFarlane S, Mather J, Ackerman T (2007) Analysis of tropical radiative heating profiles: A comparison of models and observations. J Geophys Res 112(D14): D14218. 2 of 2 | www.pnas.org/cgi/doi/10.1073/pnas.1603130113 Levermann et al.