A global assessment of actors and their roles in climate change adaptation

Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1250 na u e clima e change Analysis h ps://doi.o g/10.1038/s41558-023-01824-z A global assessmen o ac o s and hei oles in clima e change adap a ion Jan Pe zold   1 , Tom Hawxwell2, Ke s in Jan ke   3, Edua do Gonçal es G esse3, Cha lo a Mi bach1,3, Idowu Ajibade4, Su uchi Bhadwal5, Ka h yn Bowen6, Alexand a Paige Fische 7, Elphin Tom Joe8, Ch is ine J. Ki chho    9, Ka ha ine J. Mach10,11, Diana Reckien   12, Alcade C. Segnon   13,14, Chandni Singh   15, Nicola Uliba i   16, Dono an Campbell   17, Emilie C emin18, Leonie Fä be 19, G eeshma Hegde   20, Jihye Jeong21, Ab aham Ma shall Nunbogu22, Himansu Kesa i P adhan   23, Lea S. Sch öde    21, Mohammad Aminu Rahman Shah24, Pauline Reese   25, Fe dous Sul ana26, Ca los Tello27, Ji en Xu   28,29, The Global Adap a ion Mapping Ini ia i e Team* & Ma hias Ga schagen   1 An assessmen o he global p og ess in clima e change adap a ion is u gen ly needed. Despi e a ising awa eness ha adap a ion should in ol e di e se socie al ac o s and a sha ed sense o esponsibili y, li le is known abou he ypes o ac o s, such as s a e and non-s a e, and hei oles in di e en ypes o adap a ion esponses as well as in di e en egions. Based on a la ge n-s uc u ed analysis o case s udies, we show ha , al hough indi iduals o households a e he mos p ominen ac o s implemen ing adap a ion, hey a e he leas in ol ed in ins i u ional esponses, pa icula ly in he global sou h. Go e nmen s a e mos o en in ol ed in planning and ci il socie y in coo dina ing esponses. Adap a ion o indi iduals o households is documen ed especially in u al a eas, and go e nmen s in u ban a eas. O e all, unde s anding o ins i u ional, mul i-ac o and ans o ma ional adap a ion is s ill limi ed. These indings con ibu e o deba es a ound ‘social con ac s’ o adap a ion, ha is, an ag eemen on he dis ibu ion o oles and esponsibili ies, and in o m u u e adap a ion go e nance. The ealiza ion ha clima e change adap a ion is u gen has en e ed mains eam planning and policymaking, and people and ins i u ions a e adap ing 1–3 . The ecen Six h Assessmen Repo o he In e go e n- men al Panel on Clima e Change (IPCC) s esses he need o iden i y “who needs o ake wha ac ions and when in o de ha ans o ma ions un old a su icien speed and scale o mee he Pa is, SDG and o he policy goals”4. Socie ies ac oss he globe a e s uggling o ind and nego ia e an e ec i e and ai dis ibu ion o clima e ac ions be ween s a e ac o s, ci il socie y, ci izens, he p i a e sec o and o he ac o s. Wi h a g owing ealiza ion o he apidly inc easing adap a ion asks ahead o socie ies, social con ac s o clima e change adap a ion a e inc easingly deba ed locally and globally5. An impo an s a ing poin o such deba es, howe e , is he empi ical obse a ion o cu en oles in adap a ion. While clima e change adap a ion is unde s ood o be place-based, he oles and esponsibili y o ac ion ac oss ac o s, scales and di e se geog aphies a e o en unclea om bo h empi ical and no ma i e pe spec i es, ha is, ‘who is ac ing how’ and ‘who should ac how’, espec i ely5–8. In addi ion, adap a ion usually occu s in mul i-ac o se ings, whe e di e en ac o s play di e en oles. While s a e ac o s, Recei ed: 22 Decembe 2022 Accep ed: 25 Augus 2023 Published online: 12 Oc obe 2023 Check o upda es A ull lis o a ilia ions appea s a he end o he pape . *A lis o au ho s and hei a ilia ions appea s a he end o he pape .  e-mail: [email p o ec ed] Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1251 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z adap a ion17. The la e would sugges de ining oles in ela ion o, o example, agenda se ing, awa eness aising, ini ia ing policy, coo di- na ing in e ac ion, a ge se ing, s a egy making, inancing, en o ce- men and policy adjus men (see Me hods o a desc ip ion o he oles used in his analysis)7. Fo each o hese oles, di e en ypes o ac o s a e ypically associa ed wi h speci ic capaci ies, expec a ions and manda es, which, howe e , can be unclea , o e lapping o only jus eme ging. Findings om he Global Adap a ion Mapping Ini ia i e (GAMI)— he i s global sys ema ic mapping o pee - e iewed li e a u e on clima e change adap a ion—indica e a agmen a ion among ac o s and in e en ions in he con ex o clima e change adap a ion and li le e idence o ans o ma ional adap a ion, which e e s o adap a- ion wi h deep sys emic shi s1. Such shi s will only be possible wi h a clea dis ibu ion o oles and esponsibili ies. All ac o s need o come oge he he e wi h ideally a sha ed ag eemen on mu ual oles 5 . Gene - ally, households a e epo ed mo e o en in clima e change adap a ion esea ch han go e nmen ac o s, bu ela i e p ominence a ies ac oss global con ex s 1 . Addi ionally, go e nmen s end o p io i ize o he adap a ion in e en ions han businesses o ci il socie y o ganiza- ions, o en based on hei abili y o c ea e egula o y o ma ke -based adap a ion in e en ions 8 . O he e iews on ac o s in adap a ion dem- ons a e unclea di isions and di e ging pe cep ions o esponsibili- ies be ween ac o g oups and ac oss go e nance scales, which may esul in ba ie s o he e ec i e implemen a ion o esponses a he local scale 5,18,19 . O e coming such ba ie s, o example, h ough mo e polycen ic clima e go e nance sys ems ha b idge adap a ion ac ions and agendas ac oss s akeholde s and scales, is c ucial o enabling he mul i-ac o coo dina ion equi ed o ans o ma ional adap a ion 20,21 . While GAMI p o ides a ough global o e iew o documen ed ac o s and ypes o adap a ion, he e we aim o con ibu e o a deepe unde s anding o he ypes o oles speci ic ac o s play, in how a mul i-ac o collabo a ion is documen ed and how ac o – ole pa e ns di e ac oss geog aphical con ex s, including u ban and u al se le- men s, in o de o suppo u he esea ch on no ma i e social con- ac s o adap a ion and he ba ie s and d i e s o ans o ma ional o ins ance, a e ypically expec ed o ac on public in as uc u e o mode a e ins i u ional adap a ion in he in e es o socie y a la ge, p i a e ac o s usually a e no me wi h such expec a ions. They migh ac a he o hei immedia e bene i . Ye , depending on cul u al and ins i u ional a angemen s, he bounda ies be ween public and p i a e ac o s o be ween di e en le els o go e nmen can be unclea 9,10. The cons ella ion o ac o s, ins i u ional a angemen s and policy ins umen s cha ac e izing clima e change adap a ion go e nance in a gi en loca ion is highly con ex -speci ic 8,11 . Ins i u ional and ech- nological capaci ies and socioeconomic cha ac e is ics may di e conside ably be ween u ban and u al a eas, as well as be ween he global sou h and global no h1,12,13. While ci y go e nmen s wi h mo e esou ces and ins i u ional capaci ies, especially hose in he global no h, can be, in p inciple, ac i e and in e connec ed on unne s in clima e change adap a ion planning and implemen a ion, adap a- ion in many u ban a eas o he global sou h o en lack such esou ces and capaci ies, and local adap a ion o en in ol es mainly in o mal ac i i ies 14 . Ru al a eas, on he o he hand, p edominan ly in he global sou h, a e o en cha ac e ized by high deg ees o po e y, limi ed in as uc u e, a s ong ocus on ag icul u e and a deg ee o neglec by na ional policymake s. Howe e , u al communi ies and ac o s d aw on ich expe ience and local knowledge o cope wi h en i onmen al and clima ic haza ds15. When conside ing he di e se ac o s ac oss geog aphic con ex s as agen s who concei e and implemen adap a ion a he han as me e en i ies exposed o clima e change impac s16, i becomes necessa y o conside he ange o oles hese ac o s play. The e m ole he e e e s o an ac o ’s gene al posi ion o unc ion wi hin a la ge social sys em and in a ce ain p ocess, he e, clima e change adap a ion. Roles come wi h esponsibili ies in e ms o speci ic asks and du ies. On bo h le els, he e migh be di e ences be ween how ac o s a e ac ing and how hey should ac , o example, in ela ion o wha o he ac o s expec om hem. Fo he con ex s udied he e, oles can be linked o he adap a ion as well as he policy cycle. The o me includes assess- ing impac s, ulne abili y, isks and esilience; planning adap a ion; implemen ing adap a ion measu es; and moni o ing and e alua ing P i a e sec o (small and medium sized en e p ises) Ci il socie y (in e na ional; mul ina ional; na ional) In e na ional o mul ina ional go e nance ins i u ions Ci il socie y (sub-na ional; local) Iindi iduals o households O he To al numbe o obse a ions o each ole Go e nmen (na ional) Go e nmen (local) P i a e sec o (co po a ions) Go e nmen (sub-na ional) Academia Ac o ypes Ac o s and hei oles in adap ing o clima e change Awa eness aising Assessing Planning Financing o measu es Implemen ing Coo dina ing in e ac ion Moni o ing and e alua ion Highe obse ed equency Residuals o he Chi-Squa e Tes Compa ison o obse ed s. expec ed equency. Lowe obse ed equency Less han –2 Be ween 2 and –2 Mo e han 2 Mo e han 4 Less han –4 239310 277512 1,126453 116 Fig. 1 | Ac o s and hei oles in adap ing o clima e change. Resul s o he chi-squa e es calcula ing he esiduals, ha is, he di e ence be ween he obse ed and he expec ed equency o each combina ion o ac o and adap a ion ole, a e shown. Residuals below −2 and abo e 2 can be conside ed as con ibu ing mos o he signi ican esul (P < 0.001) o he chi-squa e es . Posi i e esiduals (g een) indica e a highe obse ed equency o an ac o – ole combina ion, and nega i e esiduals (pu ple) indica e a lowe - han-expec ed equency. The size o he ci cles co esponds o he alue o he esiduals. χ2 = 610.77, d. . = 70, P < 0.001. See Supplemen a y Table 2 o a de ailed o e iew o he esul s. Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1252 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z adap a ion22–25. We ask he ollowing ques ions: which oles do pa icu- la ac o s cu en ly play in clima e change adap a ion? Which ac o s in e ac o collabo a e in speci ic adap a ion p ac ices? To wha ex en can geog aphical pa e ns be iden i ied in ela ion o speci ic ac o ypes? How a e di e en clima e- ela ed esponses associa ed wi h speci ic ac o ypes? To wha ex en a e speci ic ac o ypes associa ed wi h mo e ans o ma ional o ms o adap a ion? To ind answe s o hese ques ions, we buil on GAMI’s global s ock ake o human adap a ion- ela ed esponses o clima e- ela ed change by e-coding he GAMI da abase acco ding o ac o s, oles and se lemen ypes, and by analysing he da a h ough desc ip i e s a is ics and eg ession analysis. GAMI is a spa ially explici da ase de i ed om coding empi ical s udies on clima e change conduc ed be ween 2013 and 2019 ( e . 1). A eam o 21 esea che s sys ema i- cally e-sc eened and e-coded all 1,682 a icles included in he GAMI da abase acco ding o ou speci ic inclusion c i e ia (see Me hods and Supplemen a y Fig. 1). The esul s o he emaining 1,472 a icles we e syn hesized and in e p e ed by 17 esea che s. While he GAMI da abase only includes published academic li e a u e on empi ically obse ed adap a ions, we a gue ha such analysis is ne e heless highly ele an as i allows us o examine how adap a ion oles a e ea ed in esea ch— and which blind spo s migh exis . Wi h apidly expanding adap a ion schola ship, li e a u e-based s udies such as his analysis can se e as a ele an in o ma ion sou ce o assessing ac ual adap a ion on he g ound, combined wi h a iangula ion wi h o he da ase s o add ess po en ial language bias and o he di icul - o-quan i y selec ion biases and gaps in he academic li e a u e. Ac o s and oles in adap a ion Ou indings e eal se e al pa e ns in which ac o s we e mo e likely han o he s o ake on speci ic adap a ion oles. O e all, indi iduals o households a e he mos equen ly epo ed ac o ype ( ep esen - ing 64% o coded a icles) and by a he mos documen ed ac o s o he ac ual implemen a ion o adap a ion measu es (Fig. 1). The ole o inancing is mainly associa ed wi h in e na ional o mul ina ional go e nance ins i u ions, he p i a e sec o , ci il socie y o ganiza ions and na ional go e nmen s. Planning is mo e o en done by go e nmen ac o s, wi h no pa icula pa e n disce nible a di e en go e nmen le els. Ci il socie y o ganiza ions a e impo an ac o s o coo dina ing he in e ac ion be ween a ious o he ac o s and aising awa eness. Awa eness aising is also a ele an ole associa ed wi h academia, along wi h assessing clima e impac s and moni o ing adap a ion e o s. While he e is li le epo ing on p i a e sec o ac o s ac oss all egions, hei documen ed oles a e mainly ela ed o inancing and implemen - ing adap a ion esponses. Mul i-ac o cons ella ions in adap a ion In almos a hi d o he coded a icles (404 a icles), indi iduals o households a e he sole epo ed ac o ype (Fig. 2). The mos com - mon combina ion o ac o s is indi iduals o households wi h na ional go e nmen s (79 a icles). The second mos common combina ion o ac o ypes is indi iduals o households wi h local go e nmen o sub-na ional ci il socie y (46 a icles each). The e is limi ed e idence o cases in ol ing na ional and local go e nmen s and cons ella ions wi h di e se ac o ypes ( o example, cons ella ions in ol ing go - e nance ac o s oge he wi h he p i a e sec o o ci il socie y ac o s and indi iduals o households). The mos common example o such a cons ella ion links indi iduals o households wi h local ci il socie y and he na ional go e nmen (17 a icles). The e is also li le epo ed e idence o small and medium-sized en e p ises (SMEs) collabo a - ing wi h go e nmen ac o s (se en a icles wi h SMEs and na ional go e nmen ; six a icles wi h SMEs and local go e nmen ). Al hough some examples o ac i i ies in ol e p i a e households and SMEs (16 a icles), epo s o SMEs and o he ac o ypes a e a e. Ac o s om academia mainly appea in collabo a ion wi h indi iduals o house- holds (20 a icles) and local go e nmen ( i e a icles). U ban– u al ac o pa e ns The majo i y o he publica ions (65%) epo adap a ion in u al a eas. He e, indi iduals o households a e by a he mos equen ly epo ed ac o s (47%; Fig. 3). In compa ison, go e nmen ac o s ( om local o na ional le els) make o a much smalle ac ion o he epo ed ac o s ( oge he 25%). In con as , he pic u e is almos in e se in u ban se - ings. He e, go e nmen ac o s (ac oss le els) accoun o 44% o he epo ed ac o s, while indi iduals o households con ibu e only 22%. In addi ion, na ional go e nmen s and sub-na ional ci il socie y ac o s a e epo ed mo e o en han local go e nmen s in u al con ex s, while in u ban a eas, he ocus is on local go e nmen ac o s. The e is Indi iduals o households Go e nmen (local) 253 Ci il socie y (in e na ional; mul ina ional; na ional) 203 P i a e sec o (co po a ions) 141 Go e nmen (na ional) 125 Academia 94 Ci il socie y (sub-na ional; local) 75 P i a e sec o (small and medium sized en e p ises) 63 Go e nmen (sub-na ional) 352 O he 35 Ac o ypes Mul i-ac o cons ella ions in clima e change adap a ion 883 Numbe o publica ions ea u ing each ac o ype Combina ions o c oss-ac o collabo a ion In e na ional o mul ina ional go e nance ins i u ions 102 F equency o each combina ion ype 404 5 79 46 34 20 78 9 10 14 16 27 Fig. 2 | Mul i-ac o cons ella ions in adap a ion. F equency o ac o ypes epo ed (g een ho izon al ba s), as well as how o en hey we e epo ed as single ac o ype in a s udy (single do s and ed e ical ba s) o in combina ion wi h o he ac o ypes (connec ed do s and ed e ical ba s). Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1253 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z li le di e ence be ween u ban and u al case s udies o o he ac o ypes. In ambiguous s udies— hose wi h mid-sized popula ions ha a e nei he clea ly u ban no u al o mixed-case s udies—indi iduals o households a e epo ed mos o en (26%), ollowed by na ional and local go e nmen s. Ac o s by egion The dis ibu ion o ac o s epo ed in adap a ion s udies a ies ac oss egions. Indi iduals o households a e he mos equen ly men ioned ac o s in all egions (Fig. 3). In A ica, Asia, Cen al and Sou h Ame ica, and he Small Island S a es, na ional go e nmen s a e he second mos common ac o ype. In Eu ope, na ional and local go e nmen s a e a he same a io (bo h 19%), and in No h Ame ica and Aus alasia, local go e nmen s a e men ioned sligh ly mo e o en. Howe e , when conside ing all h ee le els o go e nmen ac o s in combina ion ( ha is, local, sub-na ional and na ional le el), a clea pa e n eme ges: in highe -income egions, pa icula ly Eu ope, No h Ame ica and Aus- alasia, go e nmen ac o s ep esen he la ges g oup o epo ed adap a ion ac o s ( ha is, la ge han indi iduals/households). In con- as , indi iduals/households a e epo ed much mo e o en han all go e nmen ac o s combined in Asia and pa icula ly in A ica. Ci il socie y ac o s a e epo ed pa icula ly equen ly in Cen al and Sou h Ame ica and he Small Island S a es. In he Small Island S a es, in e - na ional o mul ina ional go e nance ins i u ions a e also epo ed equen ly (14% o he cases, he la ges ac ion ac oss all egions). The global scale u al- o-u ban di e ences be ween go e nmen ac o s and indi iduals o households (see ‘U ban– u al ac o pa e ns’) a e e en mo e p onounced when conside ing di e en wo ld egions in he global no h and global sou h. Go e nmen ac o s a e dominan ly ep esen ed in u ban a eas, pa icula ly s ongly in Aus alasia, Eu ope and No h Ame ica (see Supplemen a y Fig. 3). Fo u al a eas, egional di e ences a e smalle bu indi iduals o households a e s ill ela i ely mo e ep esen ed in A ica and Asia han in o he wo ld egions. Ac o oles by egion In e ms o ac o oles, we obse ed some commonali ies bu also di - e ences ac oss wo ld egions. Indi iduals o households a e by a he single mos epo ed ac o s o implemen ing adap a ion measu es in all wo ld egions. In all egions, go e nmen ac o s a e mainly in ol ed in planning and implemen a ion p ocesses. Especially in Eu ope, local go e nmen s a e mo e associa ed wi h planning han in o he egions. In No h Ame ica, sub-na ional go e nmen s play a g ea e ole in planning, implemen ing, moni o ing and e alua ing compa ed wi h o he egions. In con as , in A ica, sub-na ional go e nmen s a e less in ol ed in planning (Table 1 and Supplemen a y Table 3). Na u e o esponses by ac o s Excep o go e nmen s o in e na ional go e nance ins i u ions, beha iou al/cul u al esponses a e he mos common esponse o all ac o ypes. Howe e , he e a e di e ences in he ela i e associa- ion o ac o s and esponses (Fig. 4). Indi iduals and households a e mainly associa ed wi h beha iou al/cul u al and less wi h ins i u ional Indi iduals o households 378 22% 15% U ban 243 26% 19% Ambiguous 1,439 47% 14% Ru al 854 Asia 41% 15% 766 A ica 45% 15% 103 Small Island S a es 24% 20% 322 No h Ame ica 25% 15% 153 27% 17% 34% 18% 203 27% 17% 39% 11% 18% Ci il socie y (in e na ional; mul ina ional; na ional) Ci il socie y (sub-na ional; local) Go e nmen (local) Go e nmen (na ional) Go e nmen (sub-na ional) P i a e sec o (co po a ions) Academia P i a e sec o (small and medium sized en e p ises) O he In e na ional o mul ina ional go e nance ins i u ions Numbe o publica ions # Rela i e equency pe se lemen ype Rela i e equency pe geog aphy Aus alasia 267 Eu ope 21% 19% Cen al and Sou h Ame ica 14% 44% 9% 12% 39% 25% 25% 12% 30% 13% 47% 8% 41% 10% 32% 21% 11% 7% 8% 9% 7% 8% 14% 8% 9% 10% 16% 9% 19% 9% 6% 14% 6% 11% 24% 9% 7% 7% 12% 7% 9% Rela i e equency o ac o ypes as epo ed in sys ema ic li e a u e mapping 8% 8%8% 8% 7% 8% 7% 8% 6% 8% 6% 6% 6% 14% 14% 9% 9% 8% 8% 8% Fig. 3 | Regional and u ban– u al ac o pa e ns in adap a ion. Rela i e equency o ac o ypes epo ed in publica ions pe se lemen ype and egion. Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1254 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z esponses. Fo go e nmen ac o s, including in e na ional and mul- ina ional go e nance ins i u ions, ins i u ional esponses a e mos common. O e all, he e is limi ed e idence in he GAMI da abase o mo e ans o ma ional o ms o adap a ion ha include deep sys emic shi s. All ac o ypes a e mos o en associa ed wi h low-dep h adap a ions, ha is, esponses cha ac e ized by a con inua ion o exis ing p ac ices o only small changes. The analysis o how esponses o high dep h a e cha ac e ized conce ning ac o ypes, oles, esponse ypes, egions and se lemen ypes e eals ha hey a e equen ly associa ed wi h case s udies in Small Island S a es ( egion), ins i u ional adap a ion ( esponse ype), in e na ional go e nance ins i u ions (ac o ) and he coo dina ion o adap a ion ( ole). They a e less likely o be associa ed wi h s udies epo ing on beha iou al/cul u al esponses and he ole o moni o ing and e alua ing (Table 2). Discussion The analysis o clima e change adap a ion documen ed in he scien i ic li e a u e e eals impo an pa e ns ega ding ac o s and he oles hey pe o m. Indi iduals and households a e he ac o s epo ed o be implemen ing adap a ion he mos . Go e nmen ac o s, on he o he hand, a e epo ed o be p edominan ly in ol ed in planning and und - ing adap a ion. The s ong e idence o indi iduals o households as p ime ac o s o he implemen a ion o ac ual adap a ion aligns wi h ea lie esea ch concluding ha adap a ion is o en a highly localized phenomenon wi h a endency owa ds au onomous and inc emen al adap a ion26–28. This pa e n is u he suppo ed by ou inding ha indi iduals and households a e mos ly engaged wi h beha iou al/ cul u al esponses. The ela i ely low ep esen a ion o go e nmen s in he ac ual implemen a ion o adap a ion may indica e an implemen- a ion gap. Go e nmen s a e he ac o s o a i y poli ical agendas and ag eemen s o clima e ac ion ( om he global o he local le el) and could be expec ed o also ake a leading ole in he implemen a ion o ac ual adap a ion p ojec s. On he o he hand, his low ep esen a ion migh pa ly be due o he ac ha some o ms o go e nmen ac ion con ibu ing owa ds adap a ion migh no be labelled as such in he li e a u e, such as gene al po e y alle ia ion ini ia i es, social secu i y p og ammes, in as uc u e de elopmen and educa ion. The assessmen u he e ealed e idence sugges ing ha in e ms o esponse ypes, indi iduals and households play only a limi ed ole in ins i u ional adap a ion. Ins i u ional esponses may, by de ini ion, be d i en by go e nmen o ci il socie y ac o s. Howe e , in o de o acili a e he ans o ma ion o ins i u ions in a pa icipa o y and inclu- si e manne , indi iduals o households a e key ac o s, and hei ac i e engagemen has been a gued o be essen ial o he success o such p ocesses29. Mo e esea ch on he ole o hese ac o s in ins i u ional adap a ion is he e o e needed. In addi ion, ou indings yield qui e a s iking pic u e ega ding ac o s in ol ed in ecosys em-based and ech- nological/in as uc u e adap a ion. While he li e a u e s esses ha success ul ecosys em-based adap a ion elies on he ac i e in ol e- men o local communi ies 30,31 , we did no ind ha a pa icula ly s ong in ol emen o households o indi iduals, local ci il socie y ac o s o local go e nmen s has been epo ed in he li e a u e. Simila ly, we did no ind a pa icula ly s ong in ol emen o go e nmen ac o s o he p i a e sec o in echnological/in as uc u e adap a ion, despi e he ac ha bo h ac o ypes could be expec ed o play such a ole. Bo h o hese indings clea ly wa an u he esea ch and iangula ion wi h o he da a sou ces. Ou indings also iden i y ac o s hi he o unde ep esen ed in cli- ma e change adap a ion esea ch, mos no ably he p i a e sec o . This con i ms o he lines o e idence sugges ing ha he p i a e sec o lags behind in implemen ing adap a ion32–34. This lag is su p ising, gi en no only he inc easing impac s o clima e change on he p i a e sec o bu also he cen al ole o he p i a e sec o in shaping o e all de elop- men ends—and hence success ul clima e change adap a ion a la ge. Explo ing he exis ing and po en ial oles o p i a e sec o ac o s in adap a ion, pa icula ly in inancing and implemen a ion p ocesses and in con ex s whe e ma ke -based in e en ions a e p omo ed, is he e o e needed o ad ance ou unde s anding o he enable s, ba - ie s and esponsibili y shi s in clima e change adap a ion o e all 35,36 . Academia is mos p ominen ly documen ed as being in ol ed in moni- o ing and e alua ion, which is a ole ha is o e all no widely co e ed in he adap a ion ac i i ies epo ed in he academic li e a u e. Hence, he e is a gene al need o u he academic a en ion o moni o ing and e alua ion, and, in addi ion, o he capaci ies o di e en ac o ypes beyond academia o ul il his ole. A be e unde s anding o ac o oles and capaci ies o moni o ing and e alua ion can also help s eamline me hods o e alua ing he e ec i eness o implemen ed adap a ion measu es and a oid maladap a ion37. While adap a ion is gene ally conside ed a mul i-ac o p ocess4, he e is limi ed e idence o collabo a ion be ween mo e han wo di - e en ac o ypes ( o example, indi iduals o households along wi h ci il socie y and go e nmen ac o s) and, he e o e, limi ed empi ical unde s anding o how such mul i-ac o p ocesses can suppo e ec- i e adap a ion. The ela i e p ominence o local ci il socie y ac o s in coo dina ing clima e- ela ed esponses indica es how a speci ic ac o migh ind a speci ic ole in adap a ion. Mo eo e , ou indings con i m ha adap a ion in clima e change ho spo s, o example, Small Island Table 1 | Pa e ns in ac o – ole combina ions ac oss wo ld egions Region Ac o – ole combina ion Residuals Obse a ions Pe cen age o egional obse a ions A ica Go e nmen (sub-na ional) + Planning −4.09 20.16 A ica Indi iduals o households + Implemen ing 7.87 333 26.51 Asia Indi iduals o households + Implemen ing 4.27 337 22.63 Eu ope Go e nmen (local) + Planning 4.39 40 7.34 Eu ope Indi iduals o households + Implemen ing −5.76 54 9.91 No h Ame ica Go e nmen (sub-na ional) + Implemen ing 4.90 24 3.65 No h Ame ica Go e nmen (sub-na ional) + Moni o ing and e alua ing 4.51 71.07 No h Ame ica Go e nmen (sub-na ional) + Planning 4.38 20 3.04 No h Ame ica Indi iduals o households + Implemen ing −5.13 77 11.72 No h Ame ica O he + Coo dina ing in e ac ion 4.32 40.61 Resul s o he chi-squa e es calcula ing he esiduals, ha is, he di e ences be ween he obse ed and he expec ed equency o each combina ion o ac o and adap a ion ole pe egion. He e all esul s o combina ions wi h esiduals below −2 and abo e 2 a e shown. Residuals below −2 and abo e 2 can be conside ed as con ibu ing mos o he signi ican esul (P < 0.001) o he chi-squa e es . See Supplemen a y Table 3 o he ull lis o es esul s. Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1255 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z S a es, is o en associa ed wi h po en ially ans o ma ional esponses, such as eloca ion. Such kinds o esponses esul in a pa icula ly s ong need o collabo a ion because hey in ol e di e en ac o s and come wi h a high po en ial o con lic ing in e es s ha need o be media ed38–40. A he same ime, he dominance o adap a ion s udies wi h a single-ac o ocus aises he ques ion o whe he his dominance is due o a esea ch bias owa ds indi idual ac o ypes o ela es o a lack o coope a ion on he g ound. Fu u e esea ch should hus mo e explici ly explo e he ela ionships be ween ac o s and hei oles in mul i-ac o cons ella ions. Ou ac o mapping u he shows ha i is impo an o conside di e ences be ween se lemen ypes when empi ically assessing and no ma i ely deba ing he oles and esponsibili ies o ac o s in adap a ion. We ound ha he majo i y o epo ed adap a ion akes place in u al a eas, which, howe e , may o a ce ain ex en e lec a esea ch bias on adap a ion in esou ce-dependen communi ies ha ocus on households and on- he-g ound esponses o clima e- ela ed changes, o example, wi hin he esea ch on communi y-based adap a- ion41. Howe e , wi h inc easing u baniza ion and angible impac s on u ban popula ions, he esea ch ocus is al eady shi ing owa ds u ban a eas14,42. Ou inding ha go e nmen ac o s—pa icula ly local go - e nmen s—play a less p ominen ole in u al a eas han in u ban a eas may e lec ha u al a eas a e o en ou side municipal ju isdic ions, and ha ag icul u al and u al de elopmen policies a e o en cen al- ized na ionally, wi h li le egional au onomy. A he same ime, he s onge ole o go e nmen ac o s in u ban adap a ion migh be d i en by he ac ha u ban adap a ion is, o a la ge pa , conce ned wi h he adjus men o in as uc u e, much o i public. In e ms o epo ed case s udies, we ound mos u al cases si ua ed in he global sou h, ha is, in egions whe e he e ec s o clima e change a e al eady clea ly el and, hus, adap a ion esponses seem pa icula ly ele an 1,43 . In addi ion o he di e ences be ween se lemen ypes, we ound s ong egional di e ences ega ding ac o oles. O e all, go e nmen ac o s a e mo e o en documen ed in he global no h, and indi iduals o households mo e p ominen ly in he global sou h. This inding sugges s ha he e a e cu en ly di e ences in he p og ess o go e nmen -led and planned adap a ion, which highligh s he ele ance o egionally sensi i e adap a ion acking ( o example, wi hin he con ex o he global s ock ake) and con ex -speci ic deba es ega ding he dis ibu- ion o esponsibili ies o equi able and e ec i e adap a ion ac ion18. In he global pic u e, he e is limi ed e idence o ans o ma- ional adap a ion1. Fu he esea ch on ac o s and implemen a ion gaps in he con ex o clima e change adap a ion may shed ligh on he ba ie s and po en ial d i e s o ans o ma ional adap a ion. As di e en no ma i e iews o adap a ion p ocesses as well as ou - comes, a ising om di e en epis emological and disciplina y en y poin s, can lead o di e en in e p e a ions o adap a ion e ec i e- ness, u u e esea ch needs o u he explo e how di e en ac o s P i a e sec o (small and medium sized en e p ises) Ci il socie y (in e na ional; mul ina ional; na ional) In e na ional o mul ina ional go e nance ins i u ions Ci il socie y (sub-na ional; local) Indi iduals o households Beha iou al/cul u al Ecosys em-based Ins i u ional Technological/in as uc u al O he Go e nmen (na ional) Go e nmen (local) P i a e sec o (co po a ions) Go e nmen (sub-na ional) Academia Ac o ypes Ac o s and ypes o esponses in adap ing o clima e change Highe obse ed equency Residuals o he Chi-Squa e Tes Compa ison o obse ed s. expec ed equency. Lowe obse ed equency Less han –2 Be ween 2 and –2 Mo e han 2 Mo e han 4 Less han –4 To al numbe o obse a ions o each esponse 609891 772531 Fig. 4 | Ac o ypes pe esponse ype. Resul s o he chi-squa e es calcula ing he esiduals, ha is, he di e ence be ween he obse ed and he expec ed equency o each combina ion o ac o and esponse ype. Residuals below −2 and abo e 2 can be conside ed as con ibu ing mos o he signi ican esul (P < 0.001) o he chi-squa e es . Posi i e esiduals (g een) indica e a highe obse ed equency o a espec i e ac o – esponse ype combina ion, and nega i e esiduals (pu ple) indica e a lowe equency han expec ed. The size o he ci cles co esponds o he alue o he esiduals. χ2 = 87.552, d. . = 30, P < 0.001. Table 2 | Cha ac e is ics o medium- o high-dep h esponses in ela ion o he a iables ac o ype, ole, esponse ype, egion and se lemen ype O de ed eg ession coe icien s.e.m. P alue Small Island S a es 1.308 0.401 0.001 Beha iou al/cul u al −0.651 0.180 0.000 Ins i u ional 0.553 0.161 0.001 In e na ional o mul ina ional go e nance ins i u ions 0.569 0.269 0.034 Coo dina ing in e ac ion 0.392 0.185 0.034 Moni o ing and e alua ing −0.736 0.260 0.005 S a is ically signi ican (P < 0.05) coe icien s o he o de ed logis ic eg ession p edic dep h le els. Coe icien s a e he log odds o low e sus medium dep h o medium e sus high dep h when a gi en a iable exis s. s.e.m., s anda d e o o he mean. S udies ha men ion Small Island S a es, o ins ance, ha e a 3.7 (exp(1.3) = 3.7) imes highe likelihood o being associa ed wi h highe dep h le els han s udies ha do no men ion Small Island S a es. See Supplemen a y Table 5 o he ull lis o esul s. Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1256 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z ame adap a ion e ec i eness44,45. Ou indings e lec an o e all a he inc emen al implemen a ion o adap a ion, whe e indi idual and household adap a ion esponses a e mainly beha iou al, low in dep h and wi h only a limi ed connec ion o ins i u ional change, while go e nmen ac o s a e in many con ex s less p ominen , pa icula ly in implemen ing adap a ion. Such inc emen al adap a ion is impo - an o local communi ies in dealing wi h clima e isks. Howe e , ou indings sugges a lack o p econdi ions o mo e ans o ma ional adap a ion, which equi es sys emic change and mul i-ac o collabo- a ion 46 , building on b oad socie al suppo and a sha ed ag eemen on oles and esponsibili ies: a social con ac o clima e ac ion and sus ainabili y ans o ma ions. Online con en Any me hods, addi ional e e ences, Na u e Po olio epo ing sum- ma ies, sou ce da a, ex ended da a, supplemen a y in o ma ion, acknowledgemen s, pee e iew in o ma ion; de ails o au ho con i- bu ions and compe ing in e es s; and s a emen s o da a and code a ail- abili y a e a ailable a h ps://doi.o g/10.1038/s41558-023-01824-z. Re e ences 1. Be ang-Fo d, L. e al. A sys ema ic global s ock ake o e idence on human adap a ion o clima e change. Na . Clim. Change 11, 989–1000 (2021). 2. IPCC: Summa y o Policymake s. In Clima e Change 2022: Impac s, Adap a ion, and Vulne abili y (eds Pö ne , H. O. e al.) (Camb idge Uni . P ess, 2022). 3. Lesnikowski, A., Fo d, J., Biesb oek, R., Be ang-Fo d, L. & Heymann, S. J. Na ional-le el p og ess on adap a ion. Na . Clim. Change 6, 261–264 (2016). 4. A a Begum, R. e al. in Clima e Change 2022: Impac s, Adap a ion, and Vulne abili y (eds Pö ne , H. O. e al.) 121–196 (IPCC, Camb idge Uni . P ess, 2022). 5. Doshi, D. & Ga schagen, M. Assessing social con ac s o u ban adap a ion h ough social lis ening on Twi e . npj U ban Sus ain. 3, 30 (2023). 6. Mees, H. Local go e nmen s in he d i ing sea ? A compa a i e analysis o public and p i a e esponsibili ies o adap a ion o clima e change in Eu opean and No h-Ame ican ci ies. J. En i on. Plann. Policy Manag. 19, 374–390 (2017). 7. Mees, H. L. P., D iessen, P. P. J. & Runhaa , H. A. C. Explo ing he scope o public and p i a e esponsibili ies o clima e adap a ion. J. En i on. Plann. Policy Manag. 14, 305–330 (2012). 8. Uliba i, N. e al. A global assessmen o policy ools o suppo clima e adap a ion. Clim. Policy 22, 77–96 (2022). 9. Dilling, L. e al. Is adap a ion success a lawed concep ? Na . Clim. Change 9, 572–574 (2019). 10. an Nieuwaal, K., D iessen, P., Spi , T. & Te mee , K. A s a e o he a o go e nance li e a u e on adap a ion o clima e change: owa ds a esea ch agenda. IOP Con . Se . Ea h En i on. Sci. 6, 362019 (2009). 11. New, R. e al. in Clima e Change 2022: Impac s, Adap a ion, and Vulne abili y (eds Pö ne , H. O. e al.) 2539–2654 (IPCC, Camb idge Uni . P ess, 2022). 12. Mo on, J. F., Solecki, W., Dasgup a, P., Dodman, D. & Ri e a Fe e, M. in Clima e Change 2014: Impac s, Adap a ion, and Vulne abili y (eds Ba os, V. R. e al.) 153–155 (IPCC, Camb idge Uni . P ess, 2014). 13. Singh, C., Bazaz, A., Ley, D., Fo d, J. & Re i, A. Assessing he easibili y o clima e change adap a ion op ions in he wa e sec o : examples om u al and u ban landscapes. Wa e Secu . 11, 100071 (2020). 14. Adelekan, I. e al. Wha he La es Science on Impac s, Adap a ion and Vulne abili y Means o Ci ies and U ban A eas (Indian Ins i u e o Human Se lemen s, 2022); h ps://doi.o g/10.24943/ SUPSV209.2022 15. Dasgup a, P. e al. in Clima e Change 2014: Impac s, Adap a ion, and Vulne abili y (eds Ba os, V. R. e al.) 613–657 (IPCC, Camb idge Uni . P ess, 2014). 16. Eisenack, K., S ecke , R., Reckien, D. & Ho mann, E. Adap a ion o clima e change in he anspo sec o : a e iew o ac ions and ac o s. Mi ig. Adap . S a eg. Glob. Change 17, 451–469 (2012). 17. 25 Yea s o Adap a ion Unde he UNFCCC (UNCC, 2019). 18. Nalau, J., P es on, B. L. & Maloney, M. C. Is adap a ion a local esponsibili y? En i on. Sci. Policy 48, 89–98 (2015). 19. Reckien, D. & Pe ko a, E. P. Who is esponsible o clima e change adap a ion? En i on. Res. Le . 14, 014010 (2019). 20. Os om, E. Polycen ic sys ems o coping wi h collec i e ac ion and global en i onmen al change. Glob. En i on. Change 20, 550–557 (2010). 21. Zie ogel, G. e al. Ve ical in eg a ion o clima e change adap a ion in he wa e sec o : lessons om decen alisa ion in A ica and India. Reg. En i on. Change 19, 2729–2743 (2019). 22. Bie mann, F., Pa be g, P., an Assel , H. & Zelli, F. The agmen a ion o global go e nance a chi ec u es: a amewo k o analysis. Glob. En i on. Poli ics 9, 14–40 (2009). 23. Chan, S. e al. Rein igo a ing in e na ional clima e policy: a comp ehensi e amewo k o e ec i e nons a e ac ion. Glob. Policy 6, 466–473 (2015). 24. Hsu, A. & Raube , R. Di e se clima e ac o s show limi ed coo dina ion in a la ge-scale ex analysis o s a egy documen s. Commun. Ea h En i on. 2, 30 (2021). 25. an Assel , H. & Zelli, F. Connec he do s: managing he agmen a ion o global clima e go e nance. En i on. Econ. Policy S ud. 16, 137–155 (2014). 26. P es on, B. L., Mus elin, J. & Maloney, M. C. Clima e adap a ion heu is ics and he science/policy di ide. Mi ig. Adap . S a eg. Glob. Change 20, 467–497 (2015). 27. Wilson, R. S., He zige , A., Hamil on, M. & B ooks, J. S. F om inc emen al o ans o ma i e adap a ion in indi idual esponses o clima e-exace ba ed haza ds. Na . Clim. Change 10, 200–208 (2020). 28. Rahman, M. F. e al. Locally led adap a ion: p omise, pi alls, and possibili ies. Ambio h ps://doi.o g/10.1007/s13280-023-01884-7 (2023). 29. Næss, L. O., Bang, G., E iksen, S. & Ve a ne, J. Ins i u ional adap a ion o clima e change: lood esponses a he municipal le el in No way. Glob. En i on. Change 15, 125–138 (2005). 30. Genele i, D. & Za do, L. Ecosys em-based adap a ion in ci ies: an analysis o Eu opean u ban clima e adap a ion plans. Land Use Policy 50, 38–47 (2016). 31. Wamsle , C. e al. Ope a ionalizing ecosys em-based adap a ion: ha nessing ecosys em se ices o bu e communi ies agains clima e change. Ecol. Soc. 21, 31 (2016). 32. Su minski, S. P i a e-sec o adap a ion o clima e isk. Na . Clim. Change 3, 943–945 (2013). 33. Golds ein, A., Tu ne , W. R., Glads one, J. & Hole, D. G. The p i a e sec o ’s clima e change isk and adap a ion blind spo s. Na . Clim. Change 9, 18–25 (2019). 34. Lei old, R., Ga schagen, M., T an, V. & Re illa Diez, J. Flood isk educ ion and clima e change adap a ion o manu ac u ing i ms: global knowledge gaps and lessons om Ho Chi Minh Ci y. In . J. Disas e Risk Reduc . 61, 102351 (2021). 35. C ick, F., Gannon, K. E., Diop, M. & Sow, M. Enabling p i a e sec o adap a ion o clima e change in sub-Saha an A ica. Wiley In e discip. Re . Clim. Change 9, e505 (2018). 36. Gannon, K. E., C ick, F., A ela, J. & Conway, D. Wha ole o mul i-s akeholde pa ne ships in adap a ion o clima e change? Expe iences om p i a e sec o adap a ion in Kenya. Clim. Risk Manag. 32, 100319 (2021). Na u e Clima e Change | Volume 13 | No embe 2023 | 1250–1257 1257 Analysis h ps://doi.o g/10.1038/s41558-023-01824-z 37. Reckien, D. e al. Na iga ing he con inuum be ween adap a ion and maladap a ion. Na . Clim. Change 13, 907–918 (2023). 38. Ba ne , J. & McMichael, C. The e ec s o clima e change on he geog aphy and iming o human mobili y. Popul. En i on. h ps://doi.o g/10.1007/s11111-018-0295-5 (2018). 39. Nunn, P. D. & McNama a, K. E. in Dealing wi h Clima e Change on Small Islands: Towa ds E ec i e and Sus ainable Adap a ion? (eds Klöck, C. & Fink, M.) 19–44 (Gö ingen Uni . P ess, 2019). 40. Filho, W. L. e al. T ans o ma i e adap a ion as a sus ainable esponse o clima e change: insigh s om la ge-scale case s udies. Mi ig. Adap . S a eg. Glob. Change 27, 20 (2022). 41. Ki kby, P., Williams, C. & Huq, S. Communi y-based adap a ion (CBA): adding concep ual cla i y o he app oach, and es ablishing i s p inciples and challenges. Clim. De . 10, 577–589 (2018). 42. Callaghan, M. e al. Machine-lea ning-based e idence and a ibu ion mapping o 100,000 clima e impac s udies. Na . Clim. Change 11, 966–972 (2021). 43. Nalau, J. & Ve all, B. Mapping he e olu ion and cu en ends in clima e change adap a ion science. Clim. Risk Manag. 32, 100290 (2021). 44. Singh, C. e al. In e oga ing ‘e ec i eness’ in clima e change adap a ion: 11 guiding p inciples o adap a ion esea ch and p ac ice. Clim. De . 14, 650–664 (2022). 45. Doshi, D. & Ga schagen, M. Rup u es in pe cei ed solu ion spaces o adap a ion o lood isk: heu is ic insigh s om Mumbai and gene al lessons. Clim. Risk Manag. 41, 100524 (2023). 46. Piso , A. C. e al. E ec i e clima e change adap a ion means suppo ing communi y au onomy. Na . Clim. Change 12, 213–215 (2022). Publishe ’s no e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons license, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons license and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew a copy o his license, isi h p://c ea i ecommons. o g/licenses/by/4.0/. © The Au ho (s) 2023 1Depa men o Geog aphy, Ludwig-Maximilians-Uni e si ä München, München, Ge many. 2DFG-Resea ch T aining G oup ‘U ban u u e-making’, Ha enCi y Uni e si y Hambu g, Hambu g, Ge many. 3Cen e o Ea h Sys em Resea ch and Sus ainabili y, Uni e si y o Hambu g, Hambu g, Ge many. 4Depa men o En i onmen al Sciences, Emo y Uni e si y, A lan a, GA, USA. 5The Ene gy and Resou ces Ins i u e, New Delhi, India. 6Melbou ne Clima e Fu u es & Melbou ne School o Popula ion and Global Heal h, Uni e si y o Melbou ne, Melbou ne, Vic o ia, Aus alia. 7School o E n i o n men a n d S u s a i n ab i l i y, U n i e s i y o M i c hi g a n, A n n A b o , M I , U S A . 8Depa men o Ci il & En i onmen al Enginee ing, Penn S a e, Uni e si y Pa k, PA, USA. 9School o Enginee ing Design and Inno a ion, Depa men o Ci il & En i onmen al Enginee ing, Penn S a e, Uni e si y Pa k, PA, USA. 10Depa men o En i onmen al Science and Policy, Rosens iel School o Ma ine, A mosphe ic, and Ea h Science, Uni e si y o Miami, Miami, FL, USA. 11Leona d and Jayne Abess Cen e o Ecosys em Science and Policy, Uni e si y o Miami, Co al Gables, FL, USA. 12Depa men o U ban and Regional Planning and Geo-In o ma ion Managemen , Facul y o Geo-In o ma ion Science and Ea h Obse a ion, Uni e si y o Twen e, Enschede, The Ne he lands. 13In e na ional Cen e o T opical Ag icul u e (CIAT), Daka , Senegal. 14Facul y o Ag onomic Sciences, Uni e si y o Abomey-Cala i, Co onou, Republic o Benin. 15School o En i onmen and Sus ainabili y, Indian Ins i u e o Human Se lemen s, Bangalo e, India. 16Depa men o U ban Planning & Public Policy, Uni e si y o Cali o nia I ine, I ine, CA, USA. 17Depa men o Geog aphy & Geology, Uni e si y o he Wes Indies, Kings on, Jamaica. 18School o Social and En i onmen al Sus ainabili y and In e disciplina y S udies, Uni e si y o Glasgow, Glasgow, UK. 19Ins i u e o Ma ine Ecosys em and Fishe y Science, Cen e o Ea h Sys em Resea ch and Sus ainabili y, Uni e si y o Hambu g, Hambu g, Ge many. 20Maha ash a, India. 21Resea ch Uni Sus ainabili y and Clima e Risks, Cen e o Ea h Sys em Resea ch and Sus ainabili y, Uni e si y o Hambu g, Hambu g, Ge many. 22Depa men o Geog aphy and En i onmen al Managemen , Uni e si y o Wa e loo, Wa e loo, On a io, Canada. 23Helmhol z Zen um He eon, Hambu g, Ge many. 24Canadian Cen e o Clima e Change and Adap a ion, Uni e si y o P ince Edwa d Island, Cha lo e own, P ince Edwa d Island, Canada. 25Uni e si y o Copenhagen, Copenhagen, Denma k. 26Resea ch Uni Clima e Change and Secu i y, Ins i u e o Geog aphy, Cen e o Ea h Sys em Resea ch and Sus ainabili y, Uni e si y o Hambu g, Hambu g, Ge many. 27Resea ch Uni C i ical Geog aphies o Global Inequali ies, Ins i u e o Geog aphy, Cen e o Ea h Sys em Resea ch and Sus ainabili y, Uni e si y o Hambu g, Hambu g, Ge many. 28School o Social and En i onmen al Sus ainabili y, Uni e si y o Glasgow, Dum ies, UK. 29wa e @leeds, School o Geog aphy, Uni e si y o Leeds, Leeds, UK. *A ull lis o GAMI membe s appea s a h ps:// globaladap a ion.gi hub.io/.  e-mail: [email p o ec ed] The Global Adap a ion Mapping Ini ia i e Team Idowu Ajibade4, Su uchi Bhadwal5, Ka h yn Bowen6, Dono an Campbell17, Emilie C emin18, Alexand a Paige Fische 7, Ma hias Ga schagen1, Tom Hawxwell2, G eeshma Hegde20, Elphin Tom Joe8, Ch is ine J. Ki chho 9, Ka ha ine J. Mach10,11, Mohammad Aminu Rahman Shah24, Ab aham Ma shall Nunbogu22, Jan Pe zold1, Diana Reckien12, Alcade C. Segnon13,14, Chandni Singh15, Nicola Uliba i16 & Ji en Xu28,29 Na u e Clima e Change Analysis h ps://doi.o g/10.1038/s41558-023-01824-z Me hods We buil on GAMI’s sys ema ic li e a u e mapping, which p o- ides a global s ock ake o human adap a ion- ela ed esponses o clima e- ela ed changes in human sys ems 1 . Combining a no el app oach ha in eg a es sys ema ic e iewing and machine lea ning, di e se da abases we e sea ched, including he Web o Science, Scopus, PubMed and Google Schola , o assess 48,316 scien i ic documen s on adap a ion published in pee - e iewed li e a u e be ween 2013 and 2019. Wi h his me hod, only e idence on adap a ion documen ed in his body o li e a u e is included, and o he o ms o li e a u e ( o example, g ey li e a u e) a e excluded. A e he sc eening was done, based on he applied inclusion c i e ia, he da a co pus was na owed down o 1,682 publica ions ele an o human adap a ion esponses o clima e change1. Sc eening To explo e he ypes o ac o s and hei oles in clima e change adap a- ion in g ea e de ail, we conduc ed a u he sc eening and coding o he GAMI li e a u e. In he sc eening s ep, we il e ed o a icles wi h su icien empi ical in o ma ion abou he ac o s and/o he oles in he obse ed adap a ion case s udies and o a icles o which ull ex s we e a ailable. Ou o he 1,682 a icles in he GAMI li e a u e co pus, we selec ed 1,472 in his s ep (see Supplemen a y Fig. 1). Ou coding scheme included wo main ca ego ies—ac o s and oles. The de elopmen o he coding scheme was inspi ed by a quali a i e e iew app oach47,48 whe eby a leas wo independen e iewe s ead a sample o he ull ex s and gene a ed codes in esponse o he esea ch ques- ions independen ly. The au ho s hen gene a ed b oad ca ego ies in e nally and applied he eme gen amewo k o he emaining da a co pus while e u ning o p e iously coded documen s o upda e he coding o e lec newly de eloped ca ego ies. Coding The ca ego iza ion o ‘ac o s’ in clima e change adap a ion buil closely on he GAMI me hodology o p o ide he highes possible consis - ency. In he GAMI da abase, ac o s in clima e change adap a ion a e ca ego ized in o nine g oups (o ganized by sec o and le el): indi idu- als o households; ci il socie y (sub-na ional o local; in e na ional, mul ina ional o na ional); go e nmen (local; sub-na ional; na ional); in e na ional o mul ina ional go e nance ins i u ions; and p i a e sec o (SMEs, co po a ions). Du ing a es coding phase in he de elopmen o he coding scheme, we coded a sample o he GAMI li e a u e o iden i y whe he addi ional codes should be added. In se e al case s udies om he GAMI da abase, academic ins i u ions we e epo ed as ha ing an explici ole in con ibu ing o local clima e change adap a ion ini- ia i es. The e o e, we added ‘academia’ as a speci ic ac o ype. The de ini ion o an ac o o he coding is “a social en i y, ha is, a pe son o o ganiza ion, o a collec i e o pe sons and o ganiza ions, which is able o ac ”49. Mo eo e , we ocused on ‘ope a o s’ exe cising adap a- ion a he han ‘ ecep o s’ o adap a ion o ‘exposu e uni s’, while acknowledging ha hese migh o e lap16. Roles in adap a ion— he key aspec o his e iew—we e no coded in GAMI. A icles in GAMI p o ide ex eme di e si y in he deg ee o de ail p o ided o pa icula ac o s and po en ial oles. We s a ed wi h a b oad ca ego iza ion o s ages in he adap a ion p ocess 17 , on he one hand, and a pa icula ly de ailed concep ualiza ion o oles and esponsibili ies in adap a ion 7 , on he o he hand. This p ocess esul ed in a coding amewo k wi h se en oles. We dis inguished he oles o awa eness aising, assessing, planning, inancing, implemen ing, coo dina ing in e ac ion, and moni o ing and e alua ing as ele an phases in clima e change adap a ion ha may build on each o he (see Supplemen a y Fig. 2). Howe e , he espec i e oles do no always ollow he sequence shown, bu may also o e lap o a e pe o med in pa allel by di e en ac o s. Fo example, inancing and coo dina ing in e ac ions we e ound o be impo an ounda ions o any ac ion, bu can also un h oughou he phases. While he GAMI da abase p o ides in o ma ion abou egions and sec o s, including e idence om u ban and u al a eas, i does no sys ema ically dis inguish be ween hese se lemen ypes. As he oles o di e en ac o s may di e g ea ly in u ban and u al a eas, we added an addi ional code o se lemen ype, dis inguishing u ban, u al and ambiguous se lemen s50. Two independen e iewe s sc eened and coded each a icle o minimize bias, using he online pla o m Sys e 51. A hi d e iewe esol ed con lic ing codes. A e hese we e esol ed, he new da ase was me ged wi h he GAMI da ase . To add ess he ques ion ega ding egional pa e ns, we used he GAMI ca ego y ‘geog aphy’, which ca ego- izes he pape s acco ding o he IPCC egions. Rega ding di e en ypes o esponses, we used he GAMI ca ego ies, which dis inguish beha - iou al/cul u al, ecosys em-based, ins i u ional and echnological/ in as uc u al esponses. To add ess he ques ion o po en ially ans o ma ional adap- a ion, GAMI uses he ou dimensions ‘dep h’, ‘scope’, ‘speed’ and ‘limi s’. Fo he analysis in his e iew, we adop ed he GAMI coding o he ‘dep h’ ca ego y, which we conside mos ele an o he pu pose o ou e iew o ans o ma ional adap a ion. GAMI ca ego izes he dep h o obse ed adap a ions as in ol ing a low, medium o high le el o change. A high le el o dep h can se e as an indica o o ans o - ma ional adap a ion because i “ e lec (s) en i ely new p ac ices ha in ol e deep s uc u al e o m, comple e changes in mindse , majo shi s in pe cep ions o alues, and changing ins i u ional o beha - iou al no ms”1 (see Supplemen a y Table 4 o u he explana ion o he coding o ‘dep h’). Analysis All analyses we e pe o med using R s a is ical so wa e and Py hon 52 . We me ged he GAMI da abase wi h ou addi ional coding ega ding ac o s and oles. This esul ed in a da a ame ha con ains a ious ca - ego ical a iables, o which we calcula ed desc ip i e s a is ics. Fi s , we c oss- abula ed each ac o ype wi h he ole, se lemen , egion, esponse and dep h a iables. In cases whe e eco ds men ioned mo e han one ca ego y pe a iable (such as mo e han one ac o ype), he eco d was ea ed as mul iple obse a ions, wi h each gi en a new ow in he da a ame. We pe o med he chi-squa e es o independence o de e mine whe he ce ain a iable combina ions occu mo e o less equen ly han expec ed i he null hypo hesis we e co ec . We calcula ed he esiduals o iden i y which ca ego y combina ions make he g ea es con ibu ion o he chi-squa e es esul s. The es s we e pe o med o a ious a iable combina ions, such as ac o –ac o combina ions, ac o – ole combina ions and ac o – ole combina ions be ween egions. We conduc ed an o de ed logis ic eg ession using R’s MASS pack- age o unde s and a iables associa ed wi h highe dep h le els 53,54 . We speci ically included he a iables geog aphy, esponse ypes, adap a- ion oles, ac o s and se lemen ype as p edic o s o dep h le els. All ca ego ical p edic o a iables we e dummy-encoded as bina y a iables o allow o o de ed logis ic eg ession, and he esponse a iable (dep h) was encoded as an o de ed ac o con aining he le els ‘low’, ‘medium’ and ‘high’. Limi a ions The GAMI da abase only conside s e idence om esponses documen ed in pee - e iewed li e a u e and excludes me e com- mi men s, s a egies o isions. Many on- he-g ound esponses o clima e- ela ed haza ds may no be included in pee - e iewed li e a u e bu documen ed in o he o ms o li e a u e ( ha is, g ey li e a u e), o example, epo s by he p i a e sec o o ci il socie y ac o s. Simila ly, GAMI is s ongly biased owa ds English-language li e a u e due o he sea ch s ing, which excludes e idence published in o he languages