Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Metodens betydelse för kartläggning av värmesårbarhet: en jämförelse mellan PCA och additiva index i en urban miljö

Bork, Oskar LU (2026) NGEK11 20261
Department of Earth and Environmental Sciences (MGeo)
Abstract (Swedish)
Värmeböljor förväntas bli vanligare och mer intensiva till följd av klimatförändringar vilket ökar behovet av geografiska analyser av värmesårbarhet i urbana miljöer som är särskilt värmeutsatta. Denna studie undersöker hur olika metoder för konstruktion av värmesårbarhetsindex påverkar identifieringen av sårbara områden i Malmö. Två etablerade metoder jämförs: ett additivt, oviktat index och ett index baserat på Principal Component Analysis (PCA). Båda indexen konstrueras utifrån åtta sociala indikatorer kopplade till värmesårbarhet på DeSO-nivå (Malmös 209 demografiska statistikområden).
Resultaten visar att metoderna identifierar liknande övergripande geografiska mönster, med relativt hög korrelation mellan indexen. Samtidigt... (More)
Värmeböljor förväntas bli vanligare och mer intensiva till följd av klimatförändringar vilket ökar behovet av geografiska analyser av värmesårbarhet i urbana miljöer som är särskilt värmeutsatta. Denna studie undersöker hur olika metoder för konstruktion av värmesårbarhetsindex påverkar identifieringen av sårbara områden i Malmö. Två etablerade metoder jämförs: ett additivt, oviktat index och ett index baserat på Principal Component Analysis (PCA). Båda indexen konstrueras utifrån åtta sociala indikatorer kopplade till värmesårbarhet på DeSO-nivå (Malmös 209 demografiska statistikområden).
Resultaten visar att metoderna identifierar liknande övergripande geografiska mönster, med relativt hög korrelation mellan indexen. Samtidigt framträder tydliga skillnader i hur sårbarheten fördelas mellan centrala och yttre delar av Malmö. Det additiva indexet ger generellt högre och mer koncentrerade sårbarhetsvärden i centrala områden där flera indikatorer samvarierar, medan PCA-metoden ger ett mer differentierat resultat genom att reducera samvarierande indikatorer till underliggande komponenter. PCA identifierar separata dimensioner av sårbarhet kopplade till socioekonomi, hushållsstruktur och sårbara åldersgrupper. Studien visar att metodvalet har betydelse för både rangordningen av områden och vilka aspekter av sårbarhet som betonas i slutindexet. Resultaten bidrar till diskussionen om metodologiska val inom forskning om klimatrelaterad sårbarhet genom att visa att olika indexmetoder kan ge delvis olika underlag och analysdjup för klimatanpassningsarbete. Studien identifierar för- och nackdelar med båda metoder och föreslår att de kan lämpa sig bättre inom olika sammanhang, som exempelvis djupgående analys inom akademi (PCA) eller återkommande, praktisk genomförbar analys inom kommunal verksamhet (additivt index). (Less)
Abstract
Heatwaves are expected to become more frequent and intense because of climate change, increasing the need for geographical analyses of heat vulnerability, particularly in heat-exposed urban environments. This study examines how different methods for constructing heat vulnerability indices affect the identification of vulnerable areas in Malmö. Two established methods are compared: an additive, unweighted index and an index based on Principal Component Analysis (PCA). Both indices are constructed using eight social indicators of heat vulnerability at the DeSO level (209 census tracts within Malmö).
The results show that the methods identify similar overall geographical patterns, with a relatively high correlation between the indices.... (More)
Heatwaves are expected to become more frequent and intense because of climate change, increasing the need for geographical analyses of heat vulnerability, particularly in heat-exposed urban environments. This study examines how different methods for constructing heat vulnerability indices affect the identification of vulnerable areas in Malmö. Two established methods are compared: an additive, unweighted index and an index based on Principal Component Analysis (PCA). Both indices are constructed using eight social indicators of heat vulnerability at the DeSO level (209 census tracts within Malmö).
The results show that the methods identify similar overall geographical patterns, with a relatively high correlation between the indices. However, clear differences emerge in how vulnerability is distributed between central and outer parts of Malmö. The additive index generally yields higher and more concentrated values in central areas where several indicators co-vary, while the PCA method produces a more differentiated result by reducing co-varying indicators into underlying components. PCA identifies separate dimensions of vulnerability related to socioeconomic status, household structure, and vulnerable age groups.
The study shows that the choice of method affects both the ranking of areas as well as which aspects of vulnerability are emphasized in the final index. The results contribute to the discussion of methodological choices in climate-related vulnerability research by demonstrating that different index methods can provide partly different outcomes and levels of analysis for climate adaptation work. The study identifies strengths and weaknesses for both methods and suggests that they might be better suited to different contexts, such as in-depth academic analysis (PCA) or repeated and practical analysis in municipal work (additive index). (Less)
Please use this url to cite or link to this publication:
author
Bork, Oskar LU
supervisor
organization
course
NGEK11 20261
year
type
M2 - Bachelor Degree
subject
keywords
sårbarhetsindex, värme, metodjämförelse, PCA, klimatförändringar
language
Swedish
id
9234757
date added to LUP
2026-06-11 11:20:06
date last changed
2026-06-11 11:20:06
@misc{9234757,
  abstract     = {{Heatwaves are expected to become more frequent and intense because of climate change, increasing the need for geographical analyses of heat vulnerability, particularly in heat-exposed urban environments. This study examines how different methods for constructing heat vulnerability indices affect the identification of vulnerable areas in Malmö. Two established methods are compared: an additive, unweighted index and an index based on Principal Component Analysis (PCA). Both indices are constructed using eight social indicators of heat vulnerability at the DeSO level (209 census tracts within Malmö).
The results show that the methods identify similar overall geographical patterns, with a relatively high correlation between the indices. However, clear differences emerge in how vulnerability is distributed between central and outer parts of Malmö. The additive index generally yields higher and more concentrated values in central areas where several indicators co-vary, while the PCA method produces a more differentiated result by reducing co-varying indicators into underlying components. PCA identifies separate dimensions of vulnerability related to socioeconomic status, household structure, and vulnerable age groups.
The study shows that the choice of method affects both the ranking of areas as well as which aspects of vulnerability are emphasized in the final index. The results contribute to the discussion of methodological choices in climate-related vulnerability research by demonstrating that different index methods can provide partly different outcomes and levels of analysis for climate adaptation work. The study identifies strengths and weaknesses for both methods and suggests that they might be better suited to different contexts, such as in-depth academic analysis (PCA) or repeated and practical analysis in municipal work (additive index).}},
  author       = {{Bork, Oskar}},
  language     = {{swe}},
  note         = {{Student Paper}},
  title        = {{Metodens betydelse för kartläggning av värmesårbarhet: en jämförelse mellan PCA och additiva index i en urban miljö}},
  year         = {{2026}},
}