cc: 'TAR LA list' , 'TAR CLA list' , "'tar_ts@meto.gov.uk'" date: Wed, 26 Jan 2000 10:15:27 +0100 from: joos subject: Re: Technical Summary informal review to: "Griggs, Dave" Dear Dan LAs and TS writing team, Thank you very much for putting toghether such a nice technical summary. Please find below a few comments and proposals for revisions. I hope the comments are helpful. With best wishes, Fortunat General Comment: ---------------- (1) Overall, I find the text balanced, interesting, and easy to read. The structure is excellent. I very much like the leading main messages set in italic. F. Joos, Bern (2) p39: The text needs to address the consequences of stabilization scenarios, as you suggest on page 39. This is directly relevant to the UNFCCC. How do projected emissions in CO2 change when using the new generation of models? What are the potential implications of the feedbacks on this projections. What are the climatic consequences? F. Joos, Bern (3) p. 23 l 43 ff: I am somewhat unsatisfied how the present text deals with the processes responsible for the oceanic uptake of excess carbon and the natural carbon cycle. The solubility and biological pumps have been defined for an atmosphere - ocean system that is in EQUILIBRIUM and in the context of variations of the NATURAL carbon cycle, namely glacial-interglacial CO2 variations. While the pump concept is very usefull to characterize carbon cycle models and the natural carbon cycle, it may not be readily applied to a transient situation. The text under the bullet solubility pump is in my opinion not consistent with the definition of the solubility pump given in the scientific literature (Volk and Hoffert, 1985). I suggest to replace the first two bullets by the following text. Unfortunately, the new bullets are much longer and you may wish to shorten the text here and there and to improve the English. If you and the Chapter 3 authors agree on this revision, then of course several points raised below needs also to be considered in the main text of Chapter 3 (e.g. improved gas exchange rate; carbonate chemistry field studies; model comparison to WOCE, GEOSECS, TTO/SAVE results). FIRST BULLET: ------------ {\italics The solubility based mechanism for uptake of excess CO2.} Ocean uptake of CO2 emitted into the atmosphere is primerely a physicochemical process governed by three processes: (1) air-sea exchange, (2) the carbonate chemistry, and (3) the transport of excess carbon into the deep sea. (1) The net air-to-sea flux of carbon is driven by the partial pressure difference in CO2 between the atmosphere and the surface ocean. The ocean's surface layer equilibrates typically within a year to an atmospheric CO2 perturbation. Since SAR significant progress has been made in reducing the uncertainty in the gas transfer rate. (2) Gaseous CO2 dissolves in the surface ocean and reacts with water to form bicarbonate and carbonate ions. This process occurs virtually instantaneously. The chemical equilibrium between gaseous CO2 and dissolved inorganic carbon (DIC) determines together with the ocean volume primarily the ultimate {\italic uptake capacity} of the ocean for excess CO2. The system reacts such that the higher the carbon emissions into the atmosphere the higher the fraction that remains airborne. The equilibrium constants of the carbonate system are very well known both from laboratory studies and extensive field studies in the open ocean. (3) Excess DIC is transported from the surface ocean to the deep. This process occurs on timescales of decades to many centuries and is {\italic the rate limiting process} for the uptake of excess carbon. The ocean circulaton is the process most uncertain in model projections of future ocean carbon uptake. Our knowledge on the circulation are based on observations of tracers that contain a time information. Since SAR, extensive modeling studies have been carried out where modeled distribution of CFCs, and bomb- and natural-produced radiocarbon are compared to oceanic observations. The large scale features in the observed oceanic tracer fields are resolved by the models. Models have also shown to simulate well the reconstructed oceanic distribution of anthropogenic CO2. SECOND BULLET ------------- {\italics The natural carbon cycle.} The natural carbon cycle determined the pre-industrial equilibrium partitioning of carbon between the atmosphere and the ocean. The uptake of excess CO2 does not primarily depend on the natural carbon cycle. This is demonstrated by the similar results obtained with models that include the natural carbon cycle or models that include only the solubility based mechanism. The equilibrium partitioning is dependent on the non-linear interplay between ocean circulation, air-sea exchange, carbonate chemistry, the oceanic distributions of temperature, salinity, and alkalinity, and marine productivity. DIC is on average depleted in the surface ocean relative to the deep through biological and physicochemical processes. Thus, atmospheric CO2 is several hundred ppm lower for the present ocean than for an ocean that would be well mixed in DIC. The deep ocean is on average substantially cooler than the surface ocean and hence the surface ocean is depleted in DIC relative to the deep, because CO2 is more soluble in cold water. Marine ecosystem export organic carbon and calcite from surface waters to the deep where it is remineralized. This causes the surface ocean to be depleted in DIC, and enriched in alkalinity as compared to the deep. F. Joos, Bern Further comments -------------- page 9: line 13-15: Are differences in spatial coverage really a physical reason? Should the impact of limitated data resolution not be already reflected in the uncertainty range of the estimate? F. Joos, Bern p12, B6: it might be noted that trends in extrem events are notoriously difficult to detect because of their rare occurence. F. Joos, Bern p14: l. 43: 'For thousand of years before the Industrial Era GHG remained relatively constant.' CH4 concentration has changed by about a factor of two during the Holocene. F. Joos, Bern p15, l. 6-8: Does the 'very likely' (99 percent confidence) really hold also for the Alkenone and paleosol based data? Probably the 'very likely' holds only for the period covered by ice core data (400 kyr B.P.). If this is correct then the sentence might be change to consider only the last 400'000 years or modified by using a different qualifier for the pre-400 kyr period. F. Joos, Bern p15 l. 37-39: This needs some clarification. Atmospheric measurements have shown that during the last five decade the terrestrial biosphere has acted as a small sink or that the terrestrial carbon stock is close to balance. This suggests the existence of a substantial terrestrial sink process that offset carbon emission by land use, land use changes, and deforestation (estimated to be 1.8 GtC/yr+-.. during the 1980 decade). F. Joos, Bern p16: line 7: '..models can now simulate natural interannual variability of CO2'. It is suggested to delete the part related to the natural interannual variability or to modify it as suggested below. The variability in atmospheric CO2 is well established. However, how much the ocean and the terrestrial biosphere contributes to the atmospheric variability is still controversly discussed. Different reconstructions (e.g. Keeling et al, Nature 95, Francey et al., Nature 1995, Rayner et al, Tellus, 1999, Joos et al., GRL, 1999, Lee et al, Nature 1999, Feely et al, Nature 1999, LeQuere et al) yield very different results. Furthermore, the success of models to simulate the interannual variability in atm. CO2 is not always convincing. The word 'natural' is also questionable, because the variability in atm. CO2 is also modified through anthropogenic activities, e.g. biomass burning. The following sentence might be added: 'The observed interannual variability in atmospheric CO2 offers an opportunity to evaluate models of the global carbon cycle. First simulations have been successfully carried out.' p16, l. 40: typo: '.. support has some support..' F. Joos, Bern p17, l. 27: the word 'effective' might be qualified: 'radiatively effective, long-lived GHGs'. F. Joos, Bern p17, l. 35: remove word 'long-lived' here. CH4 life time of 9 years is much shorter than that of most halocarbons. F. Joos, Bern p19, l. 52: 'Figure 6.12 estimates..' English correct? F. Joos, Bern p20, l. 4ff: Land-use change: clarify that here only the impact of LUC on the albedo is discussed, but not the direct effect of GHGs emissions. F. Joos, Bern p 24, l. 16-18: Do not really understand this sentence. What do you want to say here? Uptake can be fast or slow irrespectively whether a reservoir is well mixed or not. F. Joos, Bern p25, l. 9-10: should probably read: '.. perturbations in the surface BUOYANCY balance... , sea ice formation, and the exchange of HEAT, processes that ..' Then, the text is also consistent with the last sentence of F4. p25, l. 27: 'simple models demonstrate ..' This has also been demonstrated by A/OGCMs (Frank Bryan, Manabe and Stouffer). replace 'simple models' with 'a hierarchy of models' p26, l. 12/13: 'However, not all aspects of the record have been successfully simulated' Is this not to be expected as there are random/stochastical processes in the climate system. p26, l. 31: suggest to remove: 'Nevertheless, useful information is being provided' p27, l. 40/41: 'It is necessary to .. simulate ALL aspects of the climate system.' This sentence should be modified. Probably it should read: 'It is necessary to use climate models that simulate the main processes governing the future evolution of climate.' p29, l. 5: add: '..during El Nino; this tends to decrease atmospheric CO2. On the other hand, CO2 is released from the terrestrial biota forced by changes in temperature, precipitation, droughts, and fires. p30, l. 44-46: suggest to split sentence into more sentences for clarity. p35 l.50 - p36 l. 14: clarify whether the given temperature changes are changes in the global mean or at high-latitudes. p48, l.48 Figures may be included that show the projected evolution of the atmospheric concentration of the major GHGs for the four SRES scenarios. p38 l.5 and l.6: Should the words 'sensitivity to' not be deleted? p39 F.8: Yes, stabilization scenarios are indeed important and should be included. Tab. 1: - wrong unit for CO2 in 3. line. replace pptv by ppm. - modify: 'Atmospheric lifetime of a perturbation' -- NEW FAX NUMBER; NEW FAX NUMBER; NEW FAX NUMBER; NEW FAX NUMBER; Fortunat Joos, Climate and Environmental Physics Sidlerstr. 5, CH-3012 Bern Phone: ++41(0)31 631 44 61 Fax: ++41(0)31 631 87 42 e-mail: joos@climate.unibe.ch; Internet: http://www.climate.unibe.ch/~joos/