The challenge to detect and attribute effects of climate change on human and natural systems

Dáithí A. Stone, Maximilian Auffhammer, Mark Carey, Gerrit Hansen, Christian Huggel, Wolfgang Crämer, David B. Lobell, Ulf Molau, Andrew R. Solow, Lourdes Tibig, Gary YoheClimatic 2013 · DOI

Abstract

Anthropogenic climate change has triggered impacts on natural and human systems world-wide, yet the formal scientific method of detection and attribution has been only insufficiently described. Detection and attribution of impacts of climate change is a fundamentally cross-disciplinary issue, involving concepts, terms, and standards spanning the varied requirements of the various disciplines. Key problems for current assessments include the limited availability of long-term observations, the limited knowledge on processes and mechanisms involved in changing environmental systems, and the widely different concepts applied in the scientific literature. In order to facilitate current and future assessments, this paper describes the current conceptual framework of the field and outlines a number of conceptual challenges. Based on this, it proposes workable cross-disciplinary definitions, concepts, and standards. The paper is specifically intended to serve as a baseline for continued development of a consistent cross-disciplinary framework that will facilitate integrated assessment of the detection and attribution of climate change impacts.

Introduction: The Detection and Attribution Gap

While climate scientists have become proficient at detecting human fingerprints in atmospheric variables like temperature and precipitation (the domain of IPCC Working Group I), doing the same for impacts—such as crop yields, species migration, or economic shifts—is significantly more complex.

The authors identify three primary hurdles:

  1. Confounding Drivers: Human and natural systems are influenced by numerous non-climate factors. For instance, a decline in fish stocks might be due to warming oceans or simply overfishing.
  2. Data Scarcity: Many biological and social systems lack the multi-decadal, high-quality observational records available for meteorological data.
  3. Disciplinary Silos: Ecologists, economists, and physicists use different terminologies and standards of evidence, hindering the synthesis of findings into a coherent global assessment.

Methodological Framework

This paper proposes a formal logic for the Detection and Attribution (D&A) of impacts, synthesizing existing literature and IPCC Good Practice Guidance.

Examples of drivers and their direct and subsequent impacts.

Key methodological concepts include:

  • The Behavioral Baseline: "Detection" is redefined as demonstrating that a system has changed beyond a specified baseline characterizing its behavior in the absence of climate change.
  • Single-step vs. Multi-step Attribution: Single-step uses one model to link external drivers (e.g., CO2) directly to an impact (e.g., glacier retreat). Multi-step links separate analyses—for example, one study showing emissions cause local warming, and another showing that local warming causes glacier retreat.
  • Qualitative Integration: The framework provides a pathway for incorporating qualitative research from social sciences into traditionally quantitative D&A structures.

Key Insights

  • Detection vs. Attribution: Detecting a change (e.g., earlier bird nesting) does not automatically imply attribution. Attribution requires proving climate change was a "substantial" contributor.
  • Anthropogenic vs. Observed Climate Change: Attributing an impact to observed climate change (including natural cycles) is distinct from attributing it to human-induced change.
  • Adaptation as a Confounder: In human systems, planned adaptation (like sea walls) can mask the detection of underlying climate impacts.

Results: A Unified Schematic

The paper’s primary contribution is a workable set of cross-disciplinary definitions and standards.

Schematic map of interactions between climate, natural, and human systems.

The provided schematic map (Figure 1) illustrates how human, natural, and climate systems interact through "impact chains." It formalizes the multi-step approach as a viable way to assess complex systems where end-to-end models are unavailable. Furthermore, it establishes that confidence assessments must account for the quality, coverage, and independence of evidence.

Limitations and Future Directions

Ongoing challenges include the "Ceteris Paribus" problem, where D&A assumes "all else being equal" in a non-linear world. There are also spatial and temporal mismatches between large-scale climate models and local impacts. The authors call for the development of specific protocols for defining "behavioral baselines" across disciplines to ensure future assessments are comparable.

Summary updated

Abstract

Anthropogenic climate change has triggered impacts on natural and human systems world-wide, yet the formal scientific method of detection and attribution has been only insufficiently described. Detection and attribution of impacts of climate change is a fundamentally cross-disciplinary issue, involving concepts, terms, and standards spanning the varied requirements of the various disciplines. Key problems for current assessments include the limited availability of long-term observations, the limited knowledge on processes and mechanisms involved in changing environmental systems, and the widely different concepts applied in the scientific literature. In order to facilitate current and future assessments, this paper describes the current conceptual framework of the field and outlines a number of conceptual challenges. Based on this, it proposes workable cross-disciplinary definitions, concepts, and standards. The paper is specifically intended to serve as a baseline for continued development of a consistent cross-disciplinary framework that will facilitate integrated assessment of the detection and attribution of climate change impacts.
BibTeX
@inproceedings{stone2013challenge,
  author = {Dáithí A. Stone and Maximilian Auffhammer and Mark Carey and Gerrit Hansen and Christian Huggel and Wolfgang Crämer and David B. Lobell and Ulf Molau and Andrew R. Solow and Lourdes Tibig and Gary Yohe},
  title = {{The challenge to detect and attribute effects of climate change on human and natural systems}},
  booktitle = {Climatic},
  year = {2013},
  doi = {10.1007/s10584-013-0873-6},
}
Figure