Lawrence Livermore National Laboratory September 4 1984 Dr Herbert Friedman National Research Council Commission on Physical Sciences Mathematics and Resources 2101 Constitution Avenue Washington DC 20418 Dear Dr Friedman My apologies for the delay in responding concerning the review of the report nThe Atmospheric Effects of Nuclear Explosions n I was on travel in Europe for most of August Although I have enclosed a substantial list of specific comments and suggestions mostly on newly prepared material the vast majority are simply intended to further tighten and complete the arguments rather than being critical of the thrust of the arguments My congratulations to the committee on this draft--it is greatly improved and should be issued The majority of my previous comments were completely addressed The report much more carefully and completely presents what is known and what the uncertainties are laying an excellent framework for the further research that is needed The only general comment that I would make is that the potential interactions between the various individual influences smoke dust and chemistry seem to be underplayed While little is known about what these interactions may be e g smoke or dust leading to stratospheric perturbations which influence the ozone distribution recovery time etc or the ozone reduction in the stratosphere changing its stability in a way that might allow even greater upward spreading of the smoke it would seem that the need to investigate such synergisms should receive some emphasis Again my compliments McCray to George Carrier the committee and Larry Sincerely Michael C MacCracken Atmospheric and Geophysical Sciences Division MCM nb P S You may identify these are not expected comments as mine Responses to each comment An Equal Opportunity Employer• University of California •PO Box 808 Livermore California 94550 • Telephone 415 422-1100 • Twx 910-386-8339 UCLLL LVfvlR SPECIFIC COMMENTS AND SUGGESTIONS Page 1-3 line 4 What is Moran 1982 reference Page 1-5 line 17 The range of 2 to 6% seems much too narrow Some intense fires give well less than 1% some synthetic materials give substantially more I would suggest perhaps 1-8% or at least indicate what level of confidence e g 50% 90% to apply to the 2-6% range Alternatively just indicate that 4% is your choice and probably accurate within factor of 2 to 4 Page 1-6 line 5 1 suggest adding smoke from the ensemble of fires 2 change at the beginning to after initial mixing to allow for residual mixing during first day after injection 3 clearly indicate here whether you mean uniformly in mass or in density Page 1-6 lines 15-17 You should mention that this finding assumes an unperturbed atmosphere or at least stratosphere which is unlikely given the smoke injection Page 1-6 line 23 Change heat to sunlight Net downward longwave radiation below the smoke must increase even if the smoke has no optical depth in the infrared Page 1-7 line 11 You might mesoscale scavenging alter to say early local and intermediate Page 1-10 line 8 True but there are some theoretical limits to greater than For example it is very hard to get hemispheric average temperatures e g in a 1-D model below about 235 K Page 1-10 line 21 You might add event and the relative role of solar radiation in controlling that season's climate Page 1-12 lines 26-27 Your initial baseline hemispheric versus midlatitude loading with an extinction optical depth of 3 5 will not reduce the light level at the surface by 99% everywhere or at all times of day given scattering and the diurnal cycle Page 1-13 line 19 You might change that to many of which Page 1-14 lines 1-4 You should point out that this analysis i e focusing only on the stratospheric contribution assumes an unperturbed atmosphere A larger fraction of the dust could be important if tropospheric stabilization occurs and to treat particle coagulation in the troposphere Page 1-15 lines 2-3 The quantitative results assume an unperturbed atmosphere as well as being 1-D calculations and a qualification should be added that the situation will likely be different in a perturbed atmosphere Page 1-16 line 25 Add a through e and more Page 2-2 line 5 aerosols and their evolution in time Page 2-2 line 6 precipitation and cloud cover Page 2-2 line 8 1-D models can't do meteorology as indicated in item c just temperature but they can do sensitivity studies regarding e Page 2-2 line 11 2-D models can do all that 1-D models do so they can do item c and sensitivity studies regarding e and some can look at item f The words neglect of should be changed to do not explicitly represent since the such models do attempt to represent the effects of the eddies but especially for this problem their parameterizations are highly suspect Page 2-3 line 24 Add tests in low and high latitudes Page 2-3 line 26 Add debris and thus avoided early rainout Page 2-4 line 2 Add troposphere and in mid-latitudes Page 2-4 line 6 The Erice proceedings is now published so the Knox Global scale deposition of radioactivity 1983 reference should read from a large scale exchange International Seminar on Nuclear War 3rd Session The Technical Basis for Peace held at the Ettore Majorana Centre for Scientific Culture Erice August 19-24 1983 Page 2-4 lines 12-17 types wetness etc Page 2-5 lines 1-3 sphere Also might Also need mention need to consider various soil to consider ozone issue in perturbed atmo- Page 3-2 line 23 ally Do you really need to mention Vietnam Page 3-5 line 24 Add the present estimation Is it a key Page 4-1 line 22 Given you cite a factor of 3 to 4 uncertainty in the submicron fraction shouldn't the range 10 to 24 be larger Page 4-5 line 12 Figure Health Physics 18 357-378 4 3 goes Page 5-4 lines 11-13 But such fires large hence likely emit more smoke Page 5-4 lines 20-21 Suggest necessarily argumentative deleting back at least to Peterson 1970 are also much cooler since not as all Page 5-5 lines 9-11 But small particles that are more effectively scavenged parts of the world as undo coagulate to larger ones Page 5-7 line 9 I was under the impression that Eric Kraus' report of smoke color was not black You might say dark instead Page 5-8 line 7 Add 1 humidity moisture firebreaks and topography 1 Page 5-21 line 8 The population of 750 000 000 people being in 1100 cities with populations greater than 100 000 in just the Warsaw Pact and NATO countries seems high The combined total populations of USSR Europe and US Canada is about 1 000 000 00O I thought urban fraction towns 2500 people was about 2 3 a better reference than to Turco's report would be helpful More references for the many statements on this page would be helpful Page 5-22 lines 8-16 While these substance may burn and emit particles aren't most of these essentially non-absorbing which would make them relatively unimportant especially compared to oil and natural gas emissions which might be quite important It would be helpful throughout to rank importance of sources based on the optical absorption of the smoke they produce this would play down forest and grassland fires and again emphasize importance of urban fuel loadings particularly of non-wood organics Page 5-44 lines 7-8 essentially all should be reworded as it is essentially always true in terms of particle number and thus only meaningful when referring to particle mass fractions Page 5-44 lines 16-21 A factor of 2 variation in extinction coefficient would be very important Indeed early coagulation does not seem to alter the size distribution to bring the extinction coefficient down substantially but does prepare the aerosol for later coagulation that will do so Page 5-62 line 13 Just as solar radiation could increase plume height by heating during the day strong infrared cooling at night when some fires will certainly occur could lead to lower altitudes or even plume collapse Page 5-63 line 26 Uniformly 3 by mass usually means gm gm uniformly by density usually refers to gm cm The wording here could be improved Page 5-64 line 17 Isn't scavenging below 4 km mainly of the larger parHence the air is cleaner rather ticles leaving the smaller particles than clean Page 5-66 line 20 How can one have 63 Page 5-77 line 10 no should be not ± 53% Does this mean 10-100% Page 5-81 line 5 What is the basis for saying that tropospheric average smoke optical depths of about 1 implying absorption optical depth of about 0 3-0 5 would lead to a major perturbation e g a nuclear winter Cess has run T 1 5 in the 0SU GCM at Livermore and found greatly reduced effects in July and the effect would be even less at other times of the year Page 5-81 lines 12-13 The effects can hardly be more severe than Turco et al found based on theoretical limitations to temperature change whereas they can be much less with just a few factors of 2 e g half the smoke and half as dark etc The phrasing might be altered to say that there are as many or at least as many uncertain factors that could maintain or sustain the findings of severe cooling as there are that could moderate it somewhat Page 5-1-3 line 1 How about directional shifts in the wind 1 Page 5-2-4 line 25 What does 15 min mean Is it 15 minutes worth of global scale evaporation Based on 1 m yr of rainfall and evaporation I get closer to 1 hr for 40000 Tg Page 5-2-7 line 18 We have also simulated smoke movement in a 2-D model and given that you cite Haberle as a private communication you can cite MacCracken that way 1984 NRC CRC Meeting We also have some preliminary indication of it in our model that couples the 0SU GCM to the new 3-D version of the GRANT0UR trace species transport model again not yet written up but presented at Erice in August 1984 Page 5-2-7 line 23 The kinetic energy of winds can also be returned via adiabatic compression descent which I suspect is more important Page 5-2-10 My im ression was that the average atmospheric water vapor burden was 2 5 g cm e g s 6e Sellers' book which gives almost twice the Table's number of 7 1 x 10 Tg on a global basis Somewhere your numbers are off probably a result of there being both larger area and higher mixing ratios in the tropics I would suggest clarifying the discrepancy because your global water vapor burden is much too low for example for average precipitation rate of 1 m yr it gives a water vapor lifetime of 5 days which is a factor of 2 too short Once you correct this factor of 2 a few corrections in the text are needed Page 6-13 line 26 The Luther 1983 paper is now available in the Erice proceedings The reference is Nuclear war short-term chemical and radiative effects of stratospheric injections International Seminar on Nuclear War 3rd Session The Technical Basis for Peace held at the Ettore Majorana Centre for Scientific Culture Erice August 19-24 1983 Page 6-14 Luther's study also indicated that it would be difficult to get a very large ozone hole with explosions in the 1-Mt range You might also note here that this large explosion excursion is as you said earlier not particularly likely Page 6-23 lines 15-18 lines 15-18 Is the additional attenuation significant It is not obvious that a slight vertical redistribution 2of absorption would be significant Remember that the soot about 1 g m would be mixed vertically by wind-induced motions through about 100 m depth Page 6-26 line 1 What does significant local contamination mean How would effects compare to radionuclide hazard etc A statement such as is made without some contextual framework is not very useful Page 7-1 line 15 Spectral techniques are really only used to represent horizontal transport other processes e g radiation convection etc are done on a grid Thus spectral techniques are only a method of solving equations for a carefully located grid of points Page 7-2 lines 1-4 I MacCracken reported initial calculations of this type at Erice in August 1984 We've not yet adequately validated the scavenging algorithm but we are moving material around in such away that the moving smoke then affects the radiation which affects the dynamics which controls the precipitation fields that scavenge the smoke Two particle sizes were treated which were scavenged differently but did not transform from small to large particles The reference is MacCracken M C and J J Walton RThe Effects of Interactive Transport and Scavenging of Smoke on the Calculated Temperature Change Resulting from Large Amounts of Smoke n presented at the 4th Session of the International Seminar on Nuclear War Erice Sicily 19-24 August 1984 Page 7-2 lines 6-8 The 0SU GCM of Gates includes diurnal variations and Cess Potter and Gates reported the results of smoke calculations using the model with and without diurnal variation at the Erice 1984 meeting Reference Cess R P G L Potter and W L Gates RClimatic Impact of a Nuclear Exchange Sensitivity Studies Using a General Circulation Model n presented at the 4th Session of the International Seminar on Nuclear War Erice Sicily 19-24 August 1984 Page 7-2 line 11 cannot now be included in detail in these models and others are trying to parameterize these processes We Page 7-3 line 21 Actually by the radiation perturbations induced by the particulate clouds Except for cloud microphysical modifications Page 7-4 line 19 Much of the North Atlantic Ocean the North American smoke will form plumes over Page 7-5 lines 18-19 The North Atlantic likely at that time would be well over Europe gap would indeed fill but The gaps advect too Page 7-6 line 6 The dust injection is irrelevant for how the figure is described since there was no interaction between the smoke dust and the windfield in this calculation the dust did later however affect the temperature You should also mention that this figure moves smoke using tropospheric average winds At Erice 1984 we MacCracken and Walton did a similar scenario with a 150 Tg injection with both control unperturbed and interactive winds with 3-D transport and the 0SU model The results especially' at such early times are quite similar Page 7-8 lines 4-7 Marginally correct if the absorption coefficient is a factor of 2 less then the temperature decreases are shorter in duration and the vertical distribution of absorption is quite different The statement is also marginally true only so long as all other factors e g injected mass are the same With statements such as this you could individually dismiss many factors of 2 that might accumulate Page 7-9 lines 18-21 You need to say that these radiative calculations assume annual and diurnal average daytime radiation at 30° latitude Doing a diurnal calculation would substantially increase the diurnal average solar radiation reaching the surface as calculated by Cess Page 7-10 lines 5-7 Indeed and this is why conditions cannot get much worse than TTAPS stated see comment concerning page 1-10 line 8 and page 5-81 lines 12-13 Page 7-10 lines 18-20 And very high e g 20-50 at very early times when optical depths are Page 7-10 line 26 I would disagree that 40 Tg is in the saturation regime I suspect it is right on the margin For us 15 Tg does nothing except locally under dense smoke plumes Page 7-13 line 10 Unless one knows the jargon it sounds as if rainout increases with height you might say that particle lifetime increases rapidly with altitude Page 7-15 lines 18-20 You should state perhaps in the caption to Figure 7 5 that the total optical depth includes both smoke and dust Because TTAPS didn't make that clear or Aleksandrov didn't read carefully he used the total optical depth dust plus smoke as the absorption optical depth making two errors dust smoke and then smoke is totally absorbing Try to prevent others from doing likewise Page 7-16 lines 9-11 and Figure 7 6 It would be better if you plotted temperature change i e start from 0° rather than temperature since 286 K is a rather cold starting point for the summer case that NRC is considering For example typical mid-latitude continental summer temperatures are 25-30° C rather than the 13° C starting point used by TTAPS At least this discrepancy should be mentioned Page 7-16 lines 21-22 You might also mention that the optical properties used by LLNL are within plausible bounds but are not as absorbing as the TTAPS baseline hence one cause of differences Another relates to use of a different scavenging algorithm Page 7-17 line 15-17 4 km or so Page 7-18 line 22 Assuming substantial amounts of smoke extend above Where was the threshold discussed Page 7-23 line 13 I believe Aleksandrov assumed the dust was smoke based on my latest discussions with Stenchkov but I may be wrong Page 7-24 line 3 not unreasonable given his very large injection Page 7-28 The fogs also provide heat to radiate originally derived mainly from the oceans through the condensation process Thus fogs really can act as a heat source Page 7-32 lines 14-15 Given the NRC is emphasizing the summer case why don't you show or explain the NCAR summer results which are interesting but less dramatic They are available I have a figure from them Page 7-33 lines 14-17 It is also evident in our 3-D interactive calculation MacCracken and Walton Erice 1984 and in the MacCracken 2-D calculation presented at the NRC CRC meeting in March Page 7-48 Table 1 you need to add that units are inverse seconds Page 8-6 lines 10-13 Loading the stratosphere and the troposphere would lead to quite different effects depending on whether convection is suppressed thereby disconnecting the surface and troposphere as Cess suggests Page A-6 line 18 It 1s MacCracken 1983 and the reference is Nuclear war preliminary estimates of the climatic effects of a nuclear exchange International Seminar on Nuclear War 3rd Session The Technical Basis for Peace held at the Ettore Majorana Centre for Scientific Culture Erice August 19-24 1983