National Academies Press: OpenBook

Frontiers in Decadal Climate Variability: Proceedings of a Workshop (2016)

Chapter: Overcoming Data Limitations

« Previous: The Role of External Forcing
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Overcoming Data Limitations

Understanding of decadal climate variability is limited by the relatively short length of many climate records, meaning that only a few decadal cycles may be recorded (NRC, 1995). Limited spatial coverage also presents a challenge in that many parameters and processes are not fully characterized with available observations. Several participants discussed their work filling in observational records that can inform understanding of decadal variability.

Gaining Insight from Paleoclimate Data

The study of natural climate variability, in particular to meet the need to separate out human-induced climate effects, requires studying climate variability of the past, before humans played a role. Information about past climate conditions is contained in historical records and “proxy” indicators, such as polar ice caps, tree rings, and corals.

Decadal variations in zonal wind strength can play a role in internal climate variability and consequently the rate of global temperature rise. However, past observations of wind strength and direction for the Pacific Ocean are very sparse, said Diane Thompson from Boston University, who studies coral records to help fill in the gaps in information. Her studies have focused on the period 1910-1939, when about one-third of the 20th century GMST warming occurred despite weak external forcing, suggesting that an important role by internal variability. However, there are very few wind observations for this early 20th-century warming period with which to test the role of tropical Pacific winds in this warming.

Thompson studies corals that grew just outside of a westerly facing lagoon on Tarawa, an atoll in the central Pacific Ocean. Because the lagoon is shielded from the prevailing easterly trade winds, trace metals, particularly manganese, accumulate there gradually over time. Westerly winds associated with the onset of El Niño events produce wave action that releases manganese from the enriched lagoonal sediments, which is then incorporated into coral skeletons. The coral skeletons also record the warming and freshening (due to increased rainfall) that the resulting El Niño event brings to the island.

Thompson presented a new coral record from 1890 to 2010 showing that ENSO-related westerly winds are associated with spikes in manganese (Figure 17; Thompson et al., 2015). These spikes of manganese in the coral skeleton (and thus bursts of winds from the west) were more frequent during the early 20th-century period of rapid warming and less frequent when warming leveled off in the mid-20th century. Thompson said that this wind reconstruction corroborates and extends the idea that periods of strong Pacific (easterly) trade winds (and less frequent pulses of westerly winds) are associated with cooler equatorial Pacific surface temperature and a slower rate of global warming. Conversely, periods of weaker trade winds (and more frequent pulses of westerly winds) are associated with warmer equatorial Pacific surface temperature and a faster rate of global warming. Thompson is conducting the same study in other equatorial atolls with westerly facing lagoons to replicate and extend this record and reconstruct past trade wind variability across the western tropical Pacific.

Kim Cobb from Georgia Tech presented work using oxygen isotopes in corals to reconstruct tropical Pacific SST, precipitation, and salinity over the past 1,000 years. Cobb compared the coral records to the evolution of 20th-century Pacific Decadal Variability (PDV) to help separate natural variability from potential human-caused trends in Pacific climate. Cobb

Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Image

FIGURE 17 Westerly winds associated with El Niño events (4th row = NINO3 index; 3rd row = ensemble mean zonal winds from the 20th-century reanalysis [20CR]) are correlated with spikes in coral Mn/Ca (2nd row) and also spikes in coral delta 18 oxygen (δ18O, 1st row), indicating fresher (less salty) and warmer water associated with El Niño conditions at Tarawa. SOURCE: Thompson et al. (2015).

conducts her research in the middle of the tropical Pacific on Palmyra, Fanning, and Christmas Islands, which she notes are very well positioned for study of the history of the PDV and other decadal climate phenomena (Figure 18).

The oxygen isotopic composition (δ18O) of skeletal aragonite in reef-building corals is a well-established proxy for reconstructing tropical SST and hydrological variability. Coral δ18O is inversely proportional to temperature and positively correlated to ocean salinity. Lower coral δ18O values reflect warmer and/or wetter conditions, while higher coral δ18O values reflect cooler and/or drier conditions (e. g., Corrège, 2006). Cobb said coral δ18O records from the Northern Line Islands provide 10-20 points per year of information—equivalent to monthly resolution records that can be directly compared to monthly-resolved instrumental climate records and climate model output. A suite of coral δ18O records from the Northern Line Islands is interchangeable with SSTs from these sites over the past several decades (Figure 19). Moreover, coral δ18O records from samples thrown up onto beaches across the Northern Line Islands—referred to as “fossil” corals—are interchangeable with those derived from coral cores drilled from living coral colonies (Figure 19). By combining living coral and fossil coral archives from across the Northern Line Islands, Cobb has amassed more than six centuries worth of monthly-resolved climate reconstructions from the central tropical Pacific that span various intervals of the last millennium (Figure 20).

The existing Line Islands coral δ18O reconstruction shows a late 20th-century trend toward warmer and/or fresher conditions in the central tropical Pacific that may be associated with the rise of atmospheric GHGs. However, the relative contributions of warming versus

Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Image

FIGURE 18 Map showing location of the Northern Line Islands (circled), including Palmyra (6N, 162W), Fanning (4N, 160W), and Christmas Island (2N, 157W) with respect to a regression map of Pacific SST variability. Figure modified after Di Lorenzo et al. (2015). NOTE: Coral oxygen isotopic records from the Northern Line Islands capture natural and anthropogenic decadal-scale variability (i. e., Cobb et al., 2001; Nurhati et al., 2009, 2011). SOURCE: Adapted from Kim Cobb presentation, September 3, 2015.

Image

FIGURE 19 Modern coral δ18O records from Christmas Island track very closely to SSTs (in black) and also to fossil coral δ18O records (in green). NOTES: Comparison of monthly-resolved coral δ18O records from Christmas Island (2N, 157W) with monthly instrumental SST from the gridbox containing Christmas Island (IGOSS; Reynolds et al., 2002). Records plotted include Evans et al. (1999) (blue) and Nurhati et al. (2009) (red), as well as unpublished data from the Cobb lab (Grothe et al., 2016). All coral records were drilled from living coral colonies, except for X12-D6 (green), which represents a U/Th-dated “fossil” coral. Offsets that were applied to each of the coral δ18O records prior to plotting are referenced in the figure legend, in units of per mil. SOURCE: Kim Cobb presentation, September 3, 2015.
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Image
FIGURE 20 Delta 18 oxygen isotope reconstruction pre-1000 to 2000. Cobb’s “temperature reconstruction” so far clearly shows the later 20th-century trend toward lighter oxygen isotopes, indicating warmer and fresher ocean surface water conditions. NOTES: Monthly-resolved coral oxygen isotopic (δ188O) records (black) from Palmyra Island (6N, 162W) spanning from 930 to 1998 AD, discontinuously, shown with a 10-year running average (yellow line). All data prior to 1886 originate from so-called “fossil” corals collected on ocean-facing beaches, subsequently dated by U/Th techniques (Cobb et al., 2003a). Data from 1886 to 1998 are derived from a living coral colony (Cobb et al., 2001); all other data are fossil coral data presented in Cobb et al. (2003b) and Dee et al. (in preparation). The rough time intervals of the so-called “Medieval Climate Anomaly” and the “Little Ice Age” are also shown, together with a vertical blue bar indicating the 2sigma uncertainty associated with single coral records (see spread of offsets denoted in legend for Figure 19)—in this plot only applicable to the 930-960 AD and 1550-1580 AD intervals. All other intervals are constrained by 3-8 overlapping coral records. SOURCE: Kim Cobb presentation, September 3, 2015.

freshening in driving the observed coral δ18O trend is a key question with important implications for the detection and attribution of climate changes in this area.

Cobb advocated for the continued development of additional long, high-resolution coral reconstructions from other islands in the region (i. e., nearby Christmas Island) to constrain the natural vs. anthropogenic contributions to recent trends from across the Pacific. Recognizing that most of these high-quality, high-resolution climate reconstructions will be based on the isotopic variability of water isotopes (i. e., corals, cave stalagmites, lake records, marine sediments), Cobb underscored an urgent need to understand how seawater and rainwater δ18O vary in the Pacific through both space and time, because that variability is translated directly into geologic proxies of climate variability. New research focused on bringing new empirical constraints from observational evidence of the modern system is a promising way forward, in addition to the continued investigation of water isotope-equipped simulations of past and future climate variability. The vast trove of high-quality paleoclimate records represents the best way of extending back the instrumental record of climate long enough to characterize the spatio-temporal patterns of natural decadal-scale variability, and how they differ from those signals related to anthropogenic climate change, according to Cobb.

Integrating Land Datasets, including the Arctic

The Karl et al. (2015) result that NOAA NCEI’s Huai-Min Zhang presented (see section on Challenges in Examining Climate Trends) highlights the importance of data homogenization and bias correction in observed SSTs. Matt Menne, also of NOAA NCEI, described some of the issues encountered in NOAA’s efforts to improve land surface station temperature data known as the International Surface Temperature Initiative (ISTI). These data are being used to produce a new NOAA analysis of land surface air temperature since the late 19th century.

Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Image

FIGURE 21 Comparison of number of stations and length of records (indicated by color) in NOAA’s old land dataset (top) and the recently released land dataset (bottom). SOURCE: Matt Menne presentation, September 3, 2015.

NOAA recognized the need to create a temperature dataset for land for use in reanalysis that is as robust as the International Comprehensive Ocean-Atmosphere Dataset (ICOADs). Progress to increase the dataset has been slow because there are multiple source archives to combine, which usually exist in different formats. Other challenges include devising a system for near real-time updates, managing station histories and metadata, and developing a system for documenting, tracking, and addressing errors. NOAA has already reconciled monthly and daily data and is now working on reconciling the hourly data, according to Menne.

NOAA released a new version of the dataset with many more stations and longer records (Figure 21). In general, there is better sampling of land areas, with a higher spatial density of station records. This is important, Menne said, because artificial shifts at local stations, such as station moves, instrument changes, land use changes, and time of observation changes, can be larger than the climate signal, and the impact of such changes can be quantified by comparing nearby station records. Globally today, there is oversampling at

Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Image

FIGURE 22 Observational sampling coverage for temperature in the upper 2000 m of the ocean (both Northern Hemisphere [top] and Southern Hemisphere [bottom]) has improved dramatically since the early 2000s with new Argo data (increasing coverage indicated by shift from blue to red). Such shifts introduce large uncertainties in the calculation of climate trends. SOURCE: Abraham et al. (2013).

airport locations that were previously in city centers. Therefore, although people have been worried in the past about the effects on readings from urbanization, Menne said the more obvious signal today is from ruralization of the climate record (given that airports are in rural areas), which could introduce a cooling bias.

Menne noted that data coverage in the Arctic is particularly sparse. Filling in more data from the Arctic region could alter understanding of the GMST trend over the past 15 years quite substantially. An important question is what part of the global climate signal is being missed because of inaccurate accounting for the large changes that have occurred in the data-sparse Arctic regions?

Filling in data gaps using model-data integrations and reanalysis

Formal model-data synthesis (loosely termed data assimilation1) seeks to optimally combine information contained in observations from several data streams (many that are sparse) and models that obey known conservation laws exactly. Products of this data assimilation can then be used to study climate variability over longer timescales than the strictly observational records would allow. Different techniques lead to different pitfalls in the use of these products: for example, Veronica Nieves, NASA Jet Propulsion Laboratory, presented some results that show the limitations of ocean reanalyses products in capturing the depth trend. Reanalyses tend to overestimate the slowdown and warming rates, and thus if heat starts to be transported in deeper layers of the ocean, these products will require improved analysis at depth, according to Nieves (Nieves et al., 2015).

Patrick Heimbach of the University of Texas at Austin discussed problems with developing climate hindcast2 studies using reanalysis3 products. First, reanalyses do not account for

___________________

1 Data assimilation is a cyclical procedure in which scientists compare recently collected observational data with the forecast model output. The model is then adjusted to reflect the known output before a new forecast is initiated.

2 A hindcast refers to the output from numerical simulation or prediction models of past events or history. These hindcasts can be analyzed to determine how well the model output matches the known observations of that event.

Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Image

FIGURE 23 Visual comparison of the observations from Nieves et al. (2015) (left) to the ECCO version 4 ocean state estimate (center) and other ocean reanalysis products (right) for the trend difference between 1993-2002 and 2003 onward (00s-90s). SOURCE: Patrick Heimbach presentation, September 3, 2015.

ongoing and continual spatio-temporal improvements to observing systems. For example, the large warming trend at the end of the 1970s in ERA404 is an artifact of large changes in observational coverage at the end of the 1970s. The same problem applies to the ocean and persists today (Stammer et al., 2016), as observational coverage continues to improve, for example, in measuring temperature in the deep ocean (Figure 22).

Additionally, reanalyses were actually created for weather forecasting, not climate studies, said Heimbach. In particular, so-called ocean reanalyses, like atmospheric reanalyses, do not conserve properties over time, in particular heat and freshwater (Wunsch and Heimbach, 2013). This is a problem for assessing decadal changes in climate properties that often are subtle residuals of large regional variations (Wunsch, 2016). Doing so introduces artificial heat sources or sinks that are a product of the reanalysis method and are not a measure of the actual physical properties being represented, according to Heimbach.

___________________

3 Reanalysis refers to the reprocessing of observational data spanning an historical period using a consistent modern analysis system.

4 ERA40 is a widely respected reanalysis by the European Centre for Medium-Range Weather Forecasts (ECMRWF) of global atmosphere and surface conditions from September 1957 through August 2002. ERA-Interim, a precursor to a revised extended reanalysis product to replace ERA-40, is now available here: http://apps.ecmwf.int/datasets/data/interim-full-daily/levtype=sfc/.

Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×

Heimbach has been using the ECCO (Estimating the Circulations & Climate of the Ocean) product, which he said is deliberately not called a reanalysis product, but instead an “ocean state estimate” (Wunsch and Heimbach, 2007). The ECCO design is free of artificial heat (and freshwater) sinks and sources. It places a premium on known conservation laws and uses diverse satellite and in-situ data streams forward and backward in time. Ocean heat content estimates at depths from the latest ECCO version 4 (Forget et al., 2015) are much closer to the in-situ estimates than in the various reanalysis products considered by Nieves et al. (2015), according to Heimbach (Figure 23). ECOO version 4 has also been used to show that even though global mean SST has stagnated, ocean heat content has increased fairly steadily (Figure 24). Designing, maintaining, and coping with coherent observational records of climate quality will require long-term inter-generational commitments to sustain stable observing systems (Wunsch et al., 2013).

Image

FIGURE 24 Results from ECCO version 4 confirm that whereas SST may have stagnated over the period 1992-2011, ocean heat content increase has not. NOTES: Global mean SST anomalies fluctuate between positive and negative values over the period 1992-2012 (top), while OHC anomalies (bottom) are positive over the period 2003-2012. SOURCE: Courtesy C. Piecuch, R. M. Ponte, and P. Heimbach.
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 35
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 36
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 37
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 38
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 39
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 40
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 41
Suggested Citation:"Overcoming Data Limitations." National Academies of Sciences, Engineering, and Medicine. 2016. Frontiers in Decadal Climate Variability: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/23552.
×
Page 42
Next: Toward Predictability »
Frontiers in Decadal Climate Variability: Proceedings of a Workshop Get This Book
×
Buy Paperback | $46.00 Buy Ebook | $36.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Many factors contribute to variability in Earth’s climate on a range of timescales, from seasons to decades. Natural climate variability arises from two different sources: (1) internal variability from interactions among components of the climate system, for example, between the ocean and the atmosphere, and (2) natural external forcings, such as variations in the amount of radiation from the Sun. External forcings on the climate system also arise from some human activities, such as the emission of greenhouse gases (GHGs) and aerosols. The climate that we experience is a combination of all of these factors.

Understanding climate variability on the decadal timescale is important to decision-making. Planners and policy makers want information about decadal variability in order to make decisions in a range of sectors, including for infrastructure, water resources, agriculture, and energy.

In September 2015, the National Academies of Sciences, Engineering, and Medicine convened a workshop to examine variability in Earth’s climate on decadal timescales, defined as 10 to 30 years. During the workshop, ocean and climate scientists reviewed the state of the science of decadal climate variability and its relationship to rates of human-caused global warming, and they explored opportunities for improvement in modeling and observations and assessing knowledge gaps. Frontiers in Decadal Climate Variability summarizes the presentations and discussions from the workshop.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!