CGD 2008 Profiles in Science: Dr. Caspar Ammann

Summary of achievements

Caspar Ammann's research centers around the high-resolution climate of past centuries and millennia and how this information can help to understand what elements of future climate might be predictable as well as what potential environmental and ecological impacts are to be anticipated given various story lines of climate change scenarios.

Better natural climate forcing datasets are necessary for more realistic simulations of past centuries and millennia and for a quantitatively more thorough assessment of the cause of natural variability. The significantly increased number of ice core records from the polar ice sheets offer the opportunity of more thorough statistical description of the volcanic forcing, going from a single series to a description where every eruption is provided with a probability of that event having happened in the first place and where the magnitude can be described by a distribution. Such a probabilistic formulation of the forcings is scientifically consistent with the underlying goal of quantifying the link between the forcing and climate.

Climate model studies that investigate the effects of externally forced climate are being performed. Process studies of the effects of low and high latitude volcanic eruptions as well as simulations of solar irradiance/activity changes offer insight into the models ensemble response of the climate system. In collaboration with Post-Doc David Schneider, these simulations are also serving an interdisciplinary Arctic research collaborative that attempts to synthesize the climate record gained primarily from high-resolution lake records. Collaborations across ESSL are used to identify what model configurations are necessary to capture the necessary processes. Using a suite of simulations with increasing model complexity from the standard CCSM towards a more complete representation of the climate system of WACCM (vertical extent, coupled chemistry), climate response to the repeated solar cycle and to the injection of volcanic aerosol are investigated. Other simulations with the coupled CCSM-3 investigate the cause of global patterns of climate during Medieval times as well as the transition into the Little Ice Age. No coupled GCM has so far reproduced the full structure of climate anomalies that can be found in the proxy record. In collaboration with Nicholas Graham (HRC & Scripps) and a team led by Kim Cobb (Georgia Tech), simulations are being performed that focus on the tropical circulation and its possible response to radiative forcing through mid-latitude teleconnections.

Extending the instrumental record through the development of improved high resolution climate reconstructions is crucial if we are to understand the role of forced changes in climate at regional scales. A fundamental problem in identifying what part of climate variability is externally forced arises from the large uncertainty in existing reconstruction methods and series. Caspar's collaborations with a multi-institution team of paleoclimatologists and statisticians is developing a new way of reconstructing climate using Bayesian Hierarchical Models as the framework (http://www.cgd.ucar.edu/ccr/ammann/CMG/). This allows for a more complete exploitation of the available climate record through inclusion of records with extremely different characteristics as well as explicit physical constraints. Addtionally, the community program of the Paleoclimate Reconstruction (PR) Challenge (http://www.pages.unibe.ch/science/prchallenge/index.html) is setup to test the accuracy of regional reconstructions and to guide the paleo reconstruction communities in developing more adequate forward models for their proxies. Using climate model output, a systematic intercomparison of the existing reconstruction approaches is performed and a double-blind setup will allow the community to identify where the next efforts need to be put in.

Publications

C.M. Ammann and E. Wahl, 2007: The importance of the geophysical context in statistical evaluations of climate reconstruction procedures. Climate Change, 85, 71-88 doi:10.1007/s10584-007-9276-x.

Abstract: A portion of the debate about climate reconstructions of the past millennium, and in particular about the well-known Mann-Bradley-Hughes ("MBH" 1998, 1999) reconstructions, has become disconnected from the goal of understanding natural climate variability. Here, we reflect on what can be learned from recent scientific exchanges and identify important challenges that remain to be addressed openly and productively by the community. One challenge arises from the real, underlying trend in temperatures during the instrumental period. This trend can affect regression-based reconstruction performance in cases where the calibration period does not appropriately cover the range of conditions encountered during the reconstruction. However, because it is tied to a unique spatial pattern driven by change in radiative balance, the trend cannot simply be removed in the method of climate field reconstruction used by MBH on the statistical argument of preserving degrees of freedom. More appropriately, the influence from the trend can be taken into account in some methods of significance testing. We illustrate these considerations as they apply to the MBH reconstruction and show that it remains robust back to AD 1450, and given other empirical information also back to AD 1000. However, there is now a need to move beyond hemispheric average temperatures and to focus instead on resolving climate variability at the socially more relevant regional scale.

Figure caption: Correction of MBH99: our emulation of the real world proxy-based MBH99 reconstruction containing full-period proxy PC-centering corrections and omission of the Gaspe-series during 1400-1449 (solid grey line) is compared to the original MBH99 reconstruction (solid black line).

---

E. Wahl and C.M. Ammann, 2007: Robustness of the Mann, Bradley, Hughes reconstruction of Northern Hemisphere surface temperatures: Examination of criticisms based on the nature and processing of proxy climate evidence. Climate Change, doi:10.1007/s10584-006-9105-7.

Abstract: The Mann et al. (1998) Northern Hemisphere annual temperature reconstruction over 1400-1980 is examined in light of recent criticisms concerning the nature and processing of included climate proxy data. A systematic sequence of analyses is presented that examine issues concerning the proxy evidence, utilizing both indirect analyses via exclusion of proxies and processing steps subject to criticism, and direct analyses of principal component (PC) processing methods in question. Altogether new reconstructions over 1400-1980 are developed in both the indirect and direct analyses, which demonstrate that the Mann et al. reconstruction is robust against the proxy-based criticisms addressed. In particular, reconstructed hemispheric temperatures are demonstrated to be largely unaffected by the use or non-use of PCs to summarize proxy evidence from the data-rich North American region. When proxy PCs are employed, neither the time period used to "center" the data before PC calculation nor the way the PC calculations are performed significantly affects the results, as long as the full extent of the climate information actually in the proxy data is represented by the PC time series. Clear convergence of the resulting climate reconstructions is a strong indicator for achieving this criterion. Also, recent "corrections" to the Mann et al. reconstruction that suggest 15th century temperatures could have been as high as those of the late-20th century are shown to be without statistical and climatological merit. Our examination does suggest that a slight modification to the original Mann et al. reconstruction is justifiable for the first half of the 15th century (~+0.05°), which leaves entirely unaltered the primary conclusion of Mann et al. (as well as many other reconstructions) that both the 20th century upward trend and high late-20th century hemispheric surface temperatures are anomalous over at least the last 600 years. Our results are also used to evaluate the separate criticism of reduced amplitude in the Mann et al. reconstructions over significant portions of 1400-1900, in relation to some other climate reconstructions and model-based examinations. We find that, from the perspective of the proxy data themselves, such losses probably exist, but they may be smaller than those reported in other recent work.

Figure caption: Summary of results. Panel (a) compares the Wahl-Ammann (WA) emulation of the MBH reconstruction (red) with the original (grey). Panel (b) compares the WA reconstruction (red) with an emulation of the MM03 Energy and Environment reconstruction (pink). The MM03 emulation for 1400-1449 uses the MBH 1400 proxy network as adjusted by MM03; the MM03 emulation for 1450-1980 uses the MBH 1450 proxy network as adjusted by MM03. Panel (c) compares the WA reconstruction (red) with emulations of the MM05b Energy and Environment reconstruction. The emulations directly exclude the bristlecone/foxtail pine records from calculation of PC summaries of N. American tree ring data (which are indirectly excluded by MM05a/b, cf. "Results" in text). The MM05b emulation using the 1400 proxy network is continued through 1980 (pink), as is the MM05b emulation using the 1450 proxy network (green). Panel (d) compares the WA reconstruction (red) with a reconstruction based on exclusion of the Gaspe record over 1400-1449 and use of the MM centering convention for forming PC summaries of North American tree ring data (dark magenta). Pink-coded reconstructions show validation failure according to critieria described in Section 2.3. Zero reference level in each panel is mean value for 1902-1980 instrumental data. Instrumental data in all panels are indicated as follows. Instrumental data used in calibration and verification are shown in black: annual data for full Northern Hemisphere grid over 1902-1993, and the mean of the spatially-restricted Northern Hemisphere grid over 1854-1901 (Jones and Briffa, 1992, updated). Instrumental data for 1902-2005 from Jones and Moberg (2003, updated) are also plotted, in dark blue.

---

D. Nychka and C. Ammann, 2007: The 'hockey stick' and the 1990s: a statistical perspective on reconstructing hemispheric temperatures. Tellus, doi:10.1111/j.1600-0870.2007.00270.x.

Introduction: The short instrumental record of about 100-150 yr forces us to use proxy indicators to study climate over long timescales. The climate information in these indirect data is embedded in considerable noise, and the past temperature reconstructions are therefore full of uncertainty, which blurs the understanding of the temperature evolution. To date, the characterization and quantification of uncertainty have not been a high priority in reconstruction procedures. Here we propose a new statistical methodology to explicitly account for three types of uncertainties in the reconstruction process. Via ensemble reconstruction, we directly obtain the distribution of decadal maximum as well as annual maximum. Our method is an integration of linear regression, bootstrapping and cross-validation techniques, and it (1) accounts for the effects of temporal correlation of temperature; (2) identifies the variability of the estimated statistical model and (3) adjusts the effects of potential overfitting. We apply our method to the Northern Hemisphere (NH) average temperature reconstruction. Our results indicate that the recent decadal temperature increase is rapidly overwhelming previous maxima, even with uncertainty taken into account, and the last decade is highly likely to be the warmest in the last millennium.

Figure caption: Summary of 1000 temperature ensembles: the decadal average of mean temperature over the 1000 ensembles (blue curve) with its 95% confidence region (yellow and grey band); decadal maxima from 1000 individual ensembles (purple dots) and the chance of each year corresponding to the decadal maximum (green curve scaled by the green label); the upper bound of the 95% confidence interval of the decadal maxima (dashed line) and decadal instrumental temperatures (red asterisks). The small box plots overlapped with purple dots show the distribution of decadal maxima in small groups with each group containing about 20 yr, and the leftmost big box plot shows the distribution of all the decadal maxima. The decadal average of mean temperature before 1400 (the blue curve embedded in the yellow band) and its 95% confidence region (yellow band) can be compared to the corresponding section of Fig. 3(a) in MBH99.

---

Mann, M.E., S. Rutherford, E. Wahl, and C. Ammann, 2007: Robustness of proxy-based climate field reconstruction methods Journal of Geophysical Research, 112, doi:10.1029/2006JD008272.

Abstract: Smerdon and Kaplan (hereafter SK07) have independently identified a technical issue in the original Mann et al. (2005, hereafter MRWA05) procedure that was first identified and brought to our attention by F. Zwiers and T. Lee (2006, personal communication) and since corrected (Mann et al. 2007). As we discuss below, this has no significant consequences for the results or conclusions of MRWA05 or related studies. In more recent work, Mann et al. (2007) recover the results and conclusions of MRWA05 using an implementation of the "regularized expectation maximization (RegEM)" procedure that does not suffer from the technical issue SK07 note.

Figure caption: Comparison of NH mean reconstructions using both NCAR CSM 1.4 and GKSS "Erik" simulations, based on network A, "red" proxy noise with ρ = 0.32, and two different SNR levels (0.4 and 1.0) (Table 1, experiments p-s for NCAR; experiments t-w for GKSS). (a) NCAR long (1856-1980) calibration, (b) NCAR short (1900-1980) calibration, (c) GKSS long (1856-1980) calibration, and (d) GKSS short (1900-1980) calibration. Uncertainties diagnosed from experiment s for NCAR and experiment w for GKSS.

---

Mann, M. E., S. Rutherford, E. Wahl, and C. Ammann, 2007: Comments on "Testing the Fidelity of Methods Used in Proxy-Based Reconstructions of Past Climate": The role of the standardization interval - Reply. Journal of Climate, doi:10.1175/2007JCLI1894.1

Abstract: We present results from continued investigations into the fidelity of covariance-based climate field reconstruction (CFR) approaches used in proxy-based climate reconstruction. Our experiments employ synthetic "pseudoproxy" data derived from simulations of forced climate changes over the past millennium. Using networks of these pseudoproxy data, we investigate the sensitivity of CFR performance to signal-to-noise ratios, the noise spectrum, the spatial sampling of pseudoproxy locations, the statistical representation of predictors used, and the diagnostic used to quantify reconstruction skill. Our results reinforce previous conclusions that CFR methods, correctly implemented and applied to suitable networks of proxy data, should yield reliable reconstructions of past climate histories within estimated uncertainties. Our results also demonstrate the deleterious impact of a linear detrending procedure performed recently in certain CFR studies and illustrate flaws in some previously proposed metrics of reconstruction skill.

Figure caption: Reconstruction of Northern Hemisphere mean temperature based on the RegEM CFR approach applied using "pseudoproxy" networks diagnosed from simulation of the National Center for Atmospheric Research (NCAR) Climate System Model (CSM) 1.4 simulation as in MRWA05. An 1856-1980 calibration interval is used. (a) Employing TTLS for regularization as in Mann et al. (2007), with both simulated target climate and pseudoproxy-reconstructed time series standardized over 1856-1980. (b) Results from MRWA05 (correct areal weighting has been used, as discussed in the text). Self-consistent uncertainties in the reconstructions are estimated from the unresolved residual variance during an 1856-1899 "validation" interval, based on a short (1900-1980) calibration. Actual model NH series is shown for comparison (black). All series are decadally smoothed as in MRWA05.

---

Jones P.D. and 29 colleagues: High-resolution paleoclimatology of the last millennium: a review of current status and future prospects. The Holocene.

Abstract: This review of late-Holocene palaeoclimatology represents the results from a PAGES/CLIVAR Intersection Panel meeting that took place in June 2006. The review is in three parts: the principal high-resolution proxy disciplines (trees, corals, ice cores and documentary evidence), emphasizing current issues in their use for climate reconstruction; the various approaches that have been adopted to combine multiple climate proxy records to provide estimates of past annual-to-decadal timescale Northern Hemisphere surface temperatures and other climate variables, such as large-scale circulation indices; and the forcing histories used in climate model simulations of the past millennium. We discuss the need to develop a framework through which current and new approaches to interpreting these proxy data may be rigorously assessed using pseudo-proxies derived from climate model runs, where the 'answer' is known. The article concludes with a list of recommendations.

First, more raw proxy data are required from the diverse disciplines and from more locations, as well as replication, for all proxy sources, of the basic raw measurements to improve absolute dating, and to better distinguish the proxy climate signal from noise. Second, more effort is required to improve the understanding of what individual proxies respond to, supported by more site measurements and process studies. These activities should also be mindful of the correlation structure of instrumental data, indicating which adjacent proxy records ought to be in agreement and which not. Third, large-scale climate reconstructions should be attempted using a wide variety of techniques, emphasizing those for which quantified errors can be estimated at specified timescales. Fourth, a greater use of climate model simulations is needed to guide the choice of reconstruction techniques (the pseudo-proxy concept) and possibly help determine where, given limited resources, future sampling should be concentrated.

Figure caption: Example of field reconstruction performance: composite winter temperature anomalies for Europe after 15 tropical volcanic events over the past 500 years (cf. Fischer et al., 2007). (a) Anomalies reconstructed from intrumental, documentary, and proxy records using a truncated-EOF method (multivariate PC regression, Luterbacher et al., 2004). (b) Anomalies from NCAR-CSM model (Ammann et al., 2007) over same period. (c) Pseudo-proxy reconstructed anomalies in same model context as (b), using the Wahl and Ammann (2007) emulation of the MBH98 truncated-EOF inverse regression method.

---

Batista D., P. Naveau, C. Ammann, and C. Jegat: Extracting common pulse-like forcing factors in multivariate climatic time series. Nonlin. Processes Geophys., in prep.

Abstract: To understand the full range of natural climate variability, it is important to attribute past climate variations to particular forcing factors. In this paper, our main focus is to introduce an automatic procedure to estimate the impact of strong but short-lived perturbations from large explosive volcanic eruptions on climate. An extraction algorithm that handles multivariate time series with a common but unknown forcing is presented. This statistical procedure is based on a multivariate multi-state space model and it can provide an accurate estimator of the timing and duration of the climate response to an eruption from a set of different climatic time series. It not only allows for a more objective estimation of its associated peak amplitude and the subsequent time evolution of the signal, but at the same time it provides a measure of confidence through the posterior probability for each pulse-like event. The flexibility, robustness and limitations of our approach are discussed by applying our method to simulated and real multivariate time series.

Figure caption: Three simulated time series with different trends and the common pulse-like time series displayed in Figure 2.

---

Gao C., A. Robock, and C. Ammann, 2008: Volcanic forcing of climate over the past 1500 years: An improved ice-core-based index for climate models. J. Geophys. Res., in press.

Abstract: Understanding natural causes of climate change is vital to evaluate the relative impacts of human pollution and land surface modification on climate. We have investigated one of the most important natural causes of climate change, volcanic eruptions, by using 54 ice core records from both the Arctic and Antarctica. Our recently collected suite of ice core data, more than double the number of cores ever used before, reduces errors inherent in reconstructions based on a single or small number of cores, which enables us to obtain much higher accuracy in both detection of events and quantification of the radiative effects. We extracted volcanic deposition signals from each ice core record by applying a high-pass LOESS filter to the time series and examining peaks that exceed twice the 31-year running median absolute deviation. We then studied the spatial pattern of volcanic sulfate deposition on Greenland and Antarctica, and combined this knowledge with a new understanding of stratospheric transport of volcanic aerosols to produce a forcing data set as a function of month, latitude, and altitude for the past 1500 years. We estimated the uncertainties associated with the choice of volcanic signal extraction criteria, ice-core sulfate deposition to stratospheric loading calibration factor, and the season for the eruptions without a recorded month. We forced an energy balance climate model with this new volcanic forcing data set, together with solar and anthropogenic forcing, to simulate the large scale temperature response. The results agree well with instrumental observations for the past 150 years and with proxy records for the entire period. Through better characterization of the natural causes of climate change, this new data set will lead to improved prediction of anthropogenic impacts on climate. The new data set of stratospheric sulfate injections from volcanic eruptions for the past 1500 years, as a function of latitude, altitude, and month, is available for download in a format suitable for forcing general circulation models of the climate system.