Editing IPCC:AR6/WGII/Chapter-17 (section)

==== 17.4.4.1 Overview of Knowledge Systems ====

<div id="h3-21-siblings" class="h3-siblings"></div>

AR5 emphasised the importance of knowledge systems as an enabling condition for decision-making, as did earlier ARs, all of which include a focus on the policy relevance of knowledge ( [[IPCC:Wg2:Chapter:Chapter-1#1.1.4|Section 1.1.4]] ). First introduced in IPCC reports in AR4, the term ‘knowledge system’ is used extensively in AR5 and the SRs. The discussion below follows a widely cited definition of knowledge systems as sets of interacting ‘agents, practices and institutions that organize the production, transfer and use of knowledge’ ( [[#Cornell--2013|Cornell et al., 2013]] : 61). This definition emphasises the social nature of knowledge and the importance of the link between knowledge and action, rather than presenting knowledge simply as information about past, present and future states of the world which can be of use to decision makers.

This definition of knowledge systems indicates the importance of capacity—the ability and the motivation to use knowledge for action—since capacity is an important feature which allows knowledge systems to function. Capacity is a necessary enabling condition for knowledge to be put to use in adaptation activities ( ''high confidence'' ), as shown across sectors such as water ( [[IPCC:Wg2:Chapter:Chapter-4#4.5.2|Section 4.5.2]] ), food security (Sections 5.12.3, 5.14.3), cities and settlements (Sections 6.4.2, 6.4.4) and health and well-being (Sections 7.1.3, 7.2.6), and across regions, including Africa (Sections 9.13.1, 9.14.5), Asia (Sections 10.3.6, 10.4.4) and North America ( [[IPCC:Wg2:Chapter:Chapter-14#14.4|Section 14.4.5]] ).

Some research on knowledge systems retains the earlier attention to information as a resource for decision makers. A major focus, discussed elsewhere in this chapter, has been increasing the precision about the certainty, likelihood and confidence with which certain statements are made in relation to underlying evidence (see Cross-Chapter Box DEEP in this Chapter). This topic, which was first introduced in AR4, advanced significantly in AR5 ( [[#Mach--2017|Mach et al., 2017]] ).

In addition to these characteristics of information, the social and organisational aspects of knowledge systems have also been the subject of recent research. One strand of this discussion emphasises the distinctiveness of different knowledge systems, often focusing on three types of knowledge: scientific, Indigenous and local, and the latter are two sometimes grouped as ‘traditional’ knowledge (see Cross-Chapter Box INDIG in Chapter 18). This strand emphasises the specific forms of knowledge production and circulation in each type. Another strand of discussion emphasises the networks of interactions between different groups. This strand follows the influential ‘Knowledge systems for sustainable development’ ( [[#Cash--2003|Cash et al., 2003]] ), which was cited in Chapters 2, 7 and 8 in WGII AR5; [[#Cash--2003|Cash et al. (2003)]] emphasise the usability and acceptability of scientific knowledge, and highlight the relations between knowledge producers and users. The discussion in [[#17.4.4|Section 17.4.4]] on knowledge as an enabling factor integrates these two strands of discussion of knowledge systems.

It was well established in AR5 and SRs that a component of knowledge systems for good climate decision-making is the production of ‘information on climate, its impacts, potential risks, and vulnerability’ which can ‘be integrated into an existing or proposed decision-making context’ ( [[#Jones--2014|Jones et al., 2014]] : 200). Also important are two other components of knowledge: of response options and knowledge of other enabling conditions, particularly governance and finance, which were mentioned less frequently and more indirectly in AR5 and SR1.5, SROCC and SRCCL. Decision makers assess the feasibility of different alternatives (see Cross-Chapter Box FEASIB) and develop strategies for the implementation and modification of the alternative, requiring a level of knowledge of the governance, policy and finance landscapes at national ( [[#Tanner--2019|Tanner et al., 2019]] ; [[#Lopes--2020|Lopes et al., 2020]] ; [[#Roberts--2020|Roberts et al., 2020]] ) and international scales ( [[#Woodruff--2018|Woodruff, 2018]] ).

Examples of the importance of these other two components—knowledge of response options and knowledge of enabling conditions—are provided by networks of cities, including internal institutional networks ( [[#Aylett--2015|Aylett, 2015]] ), intermunicipal networks (e.g., those supported by Local Governments for Sustainability [ICLEI] and the international United Cities and Local Governments [UCLG] network), transnational municipal networks (e.g., 100 Resilient Cities, Asian Cities Climate Change Resilience Network [ACCCRN]) and city-to-city regional transdisciplinary learning networks ( [[#Ndebele-Murisa--2020|Ndebele-Murisa et al., 2020]] ). These networks generate and exchange knowledge which can be critical to decision makers for understanding and evaluating the feasibility of different response options, identifying synergies across sectors and mainstreaming adaptation to climate change ( [[#Haupt--2020|Haupt et al., 2020]] ). However, the question of how to finance such network activities remains under-studied ( [[#Bracking--2021|Bracking, 2021]] ; See Box 17.3).

In addition to these general considerations of knowledge systems, research since AR5 has contributed to the understanding of specific types of knowledge. Scientific knowledge is thoroughly discussed in Chapter 1, especially in [[IPCC:Wg2:Chapter:Chapter-1#1.3|Section 1.3]] ‘Understanding and Evaluating Climate Risk’, which shows recent advances in the well-established IPCC categories of observation of past conditions and model-based projections of future conditions. We add here a consideration of a new area within scientific knowledge, artificial intelligence, which offers new methods for producing information that can be incorporated into knowledge systems.

Applying artificial intelligence (AI) to climate change is predominantly in the area of climate modelling and forecasting, inclusive of weather extremes ( [[#Monteleoni--2013|Monteleoni et al., 2013]] ; [[#Jones--2017|Jones, 2017]] ; [[#Huntingford--2019|Huntingford et al., 2019]] ). Recent efforts conceptualise the potential uses of AI for mitigation and adaptation ( [[#Rolnick--2019|Rolnick et al., 2019]] ; Cheong et al., 2020b) in addition to forecasting ( [[#Rolnick--2019|Rolnick et al., 2019]] ; [[#Chattopadhyay--2020|Chattopadhyay et al., 2020]] ; Cheong et al., 2020b; [[#Prabhat--2021|Prabhat et al., 2021]] ). There are very few cases to assess AI applications in these domains given that AI is a new field for climate change impact and adaptation. To this date, sectoral applications of AI relevant to climate change adaptation and risk reduction mainly have advanced in the areas of crop yields, early-warning systems and water management.

These sectoral advances using AI employ various learning techniques inclusive of supervised and unsupervised learning, multi-modal learning and transfer learning techniques to generate more accurate predictions than afforded by traditional climate projection methods (Cheong et al., 2020b; [[#Camps-Valls--2021|Camps-Valls et al., 2021]] ). AI applications use finer-resolution data such as sub-daily weather-related data, remote and wearable sensor data, text data and real-time survey data. They are fed into neural networks and semi-/unsupervised learning to configure detailed and more precise predictions of climate change impact on crop yields ( [[#Crane-Droesch--2018|Crane-Droesch, 2018]] ), early warning ( [[#Moon--2019|Moon et al., 2019]] ), impact of extreme heat on older adults ( [[#Cheong--2020a|Cheong et al., 2020a]] ), poverty in Africa ( [[#Oshri--2018|Oshri et al., 2018]] ) and multi-scale water management combining blockchain technology with remote water sensors ( [[#Lin--2018|Lin et al., 2018]] ).

Indigenous knowledge and local knowledge are thoroughly covered in SROCC ( [[#Abram--2019|Abram et al., 2019]] ; [[#IPCC--2019c|IPCC, 2019c]] ; [[#IPCC--2019d|IPCC, 2019d]] ) and in [[IPCC:Wg2:Chapter:Chapter-1#1.3.3|Section 1.3.3]] . We here add relevant points to decision-making, and an additional form of knowledge, practitioner knowledge.

Indigenous knowledge and local knowledge are gaining recognition at multiple scales ( [[#Kleiche-Dray--2016|Kleiche-Dray and Waast, 2016]] ; [[#David-Chavez--2018|David-Chavez and Gavin, 2018]] ; [[#Nakashima--2018|Nakashima et al., 2018]] ). Of note is their association with ecosystem-based adaptations, showcasing the long-term place-based knowledge of Indigenous Peoples ( [[#Johnson--2015|Johnson et al., 2015]] ; [[#Walshe--2016|Walshe and Argumedo, 2016]] ; [[#Carter--2019|Carter, 2019]] ; [[#Mazzocchi--2020|Mazzocchi, 2020]] ). These knowledges and practices can be an important enabling condition in decision-making processes, complementing scientific information by identifying impacts ( [[#Fernández-Llamazares--2017|Fernández-Llamazares et al., 2017]] ; [[#Katz--2020|Katz et al., 2020]] ), emphasising values to consider ( [[#Huambachano--2018|Huambachano, 2018]] ), offering solutions ( [[#Chanza--2016|Chanza and de Wit, 2016]] ; [[#Cuaton--2020|Cuaton and Su, 2020]] ; [[#Orlove--2020|Orlove et al., 2020]] ), guiding land use and resource management ( [[#Brondízio--2021|Brondízio et al., 2021]] ) and filling gaps in scientific knowledge ( [[#Hiwasaki--2014|Hiwasaki et al., 2014]] ; [[#Audefroy--2017|Audefroy and Sánchez, 2017]] ; [[#Makondo--2018|Makondo and Thomas, 2018]] ; [[#Son--2019|Son et al., 2019]] ; [[#Latulippe--2020|Latulippe and Klenk, 2020]] ; Wheeler et al., 2020).

Practitioner knowledge—the pragmatic, practice-based knowledge that comes from the regular exercise of craft or professional work—was also acknowledged briefly in AR5 ( [[#Jones--2014|Jones et al., 2014]] ) and treated significantly in SROCC ( [[#Abram--2019|Abram et al., 2019]] ). Practitioner knowledge resembles local knowledge in that it is acquired through participation in activities, and yet it differs from local knowledge, which is often place-based and tied directly to specific landscapes and communities. Local knowledge typically covers a variety of environmental domains. Practitioner knowledge may be shared with people in different locations and is often more focused on a narrower set of work activities. Recent calls have recommended bringing practitioners more fully into the IPCC assessment process, to promote more effective decision-making ( [[#Howarth--2018|Howarth et al., 2018]] ).

Practitioner knowledge makes significant contributions to decision-making by broadening the range of alternatives which are considered and by bringing in understandings of systems to the selection and implementation of alternatives. Such knowledge is applicable to a large number of domains, including biodiversity management ( [[#Tengö--2014|Tengö et al., 2014]] ; [[#Rathwell--2015|Rathwell et al., 2015]] ), and natural hazard risk management in urban settings, as reported in Denmark ( [[#Madsen--2019|Madsen et al., 2019]] ), the USA ( [[#Matsler--2019|Matsler, 2019]] ), Canada ( [[#Yumagulova--2019|Yumagulova and Vertinsky, 2019]] ), Mexico ( [[#Aguilar-Barajas--2019|Aguilar-Barajas et al., 2019]] ) and the Caribbean ( [[#Ramsey--2019|Ramsey et al., 2019]] ). Other contexts, all at regional scales, include watershed management in Peru ( [[#Ostovar--2019|Ostovar, 2019]] ), livestock management in Finland ( [[#Rasmus--2020|Rasmus et al., 2020]] ), agricultural adaptation in a context of water scarcity in Iran ( [[#Zarei--2020|Zarei et al., 2020]] ) and the water–energy nexus in the USA ( [[#Gim--2019|Gim et al., 2019]] ).

Literature indicates the importance of effective governance for promoting integration of local and practitioner knowledge with scientific knowledge ( ''high confidence'' ). This integration is most extensive and promotes a wider consideration of alternatives, where governance arrangements promote ongoing exchanges of information and discussion of solutions, whether through formal mechanisms such as regional committees ( [[#Gim--2019|Gim et al., 2019]] ; [[#Ostovar--2019|Ostovar, 2019]] ; [[#Rasmus--2020|Rasmus et al., 2020]] ; [[#Zarei--2020|Zarei et al., 2020]] ) or informal mechanisms such as personal networks and local discussion groups ( [[#Madsen--2019|Madsen et al., 2019]] ; [[#Yumagulova--2019|Yumagulova and Vertinsky, 2019]] ). Where such arrangements are absent, practitioner knowledge is side-lined from the formulation and implementation of decisions ( [[#Aguilar-Barajas--2019|Aguilar-Barajas et al., 2019]] ; [[#Matsler--2019|Matsler, 2019]] ; [[#Ramsey--2019|Ramsey et al., 2019]] ).

<div id="17.4.4.2" class="h3-container"></div>

<span id="co-production-and-other-composite-knowledge-systems"></span>