Modeling 08: Vulnerability Index and SWOT Synthesis

Objective

By the end of this tutorial you will be able to:

Compose a Composite Vulnerability Index (CVI) following the IPCC AR6 three-pillar structure — climate Exposure, population Sensitivity, and Adaptive Capacity — from health-climate, socioeconomic, and health infrastructure indicators, using sus_mod_vulnerability_index();
Interpret the vulnerability ranking across municipalities and measure territorial inequality with the Gini Coefficient and the Lorenz curve;
Synthesize the results into a SWOT analysis (Strengths, Weaknesses, Opportunities and Threats) oriented toward health-climate public policy, using sus_mod_swot();
Visualize and communicate the results with the paired plotting functions sus_mod_plot_vulnerability() (ranking, pillars, Lorenz) and sus_mod_plot_swot() (radar, 2×2 matrix, bars).

This is the eighth module in the Modeling series of climasus4r. The main inputs are the climasus_af object from Module M2 (mod_af.rds) and the climasus_dlnm object from Module M1 (mod_dlnm_fit.rds). The main output is the climasus_vi object saved as mod_vi.rds for reuse in final reports and risk communication.

Scientific background

Climate vulnerability according to the IPCC

The IPCC’s Fifth Assessment Report (AR5, 2014) established the three-pillar framework that defines vulnerability — later refined in AR6 (2022) — as a function of three interdependent components:

Exposure: the degree to which a system or population is affected by climate stressors (extreme temperature, heat waves, days above the 90th percentile of heat);
Sensitivity: the level of response of a system to the stressor, conditioned by intrinsic factors such as the proportion of elderly people, prevalence of chronic disease, and the dose-response relationship estimated by DLNM models;
Adaptive Capacity: the set of institutional, economic, and infrastructural resources that allow a system to modulate its impacts (human development index, treated water coverage, density of primary health care units).

The composite index formula, originally proposed by Turner et al. (2003) and adopted in the social vulnerability literature (Cutter, Boruff & Shirley, 2003; Benmarhnia et al., 2015), inverts the Adaptive Capacity pillar before combination, since higher capacity means lower vulnerability:

$VI_i = \frac{w_E \cdot E_i + w_S \cdot S_i + w_{AC} \cdot (1 - AC_i)}{w_E + w_S + w_{AC}}$

where $E_i$ , $S_i$ , and $AC_i$ are the weighted averages of the normalized indicators (min-max or z-score) of municipality $i$ for each pillar, and $w_E$ , $w_S$ , $w_{AC}$ are the inter-pillar weights. Missing indicators are handled through automatic reweighting.

Min-max and z-score normalization

Min-max normalization projects each indicator $x_{ij}$ onto the interval $[0, 1]$ :

$\tilde{x}_{ij} = \frac{x_{ij} - \min_j(x_{ij})}{\max_j(x_{ij}) - \min_j(x_{ij})}$

Z-score normalization standardizes by mean and standard deviation, preserving the original distribution — useful when indicators have very heterogeneous scales, but it requires a back-transformation for direct interpretation.

Territorial inequality and the Lorenz Curve

The Lorenz Curve is the classic instrument for visualizing the concentration of any resource or burden across an ordered population. In the context of climate risk, the curve accumulates the proportion of total CVI (y-axis) against the proportion of municipalities (x-axis), from least to most vulnerable. The larger the area between the curve and the diagonal of perfect equality, the higher the Gini Coefficient — which ranges from 0 (perfect equality) to 1 (all risk concentrated in a single municipality).

Climate risk concentration analyses have been applied at multi-city scale by Gasparrini et al. (2017) and in national socio-environmental vulnerability studies in Brazil by Confalonieri et al. (2009) and Marengo et al. (2022).

SWOT analysis oriented toward health-climate policy

The SWOT framework (Andrews, 1971), widely used in strategic planning, has been adapted to climate-health risk management. In this approach:

Strengths (S) represent existing protective factors — high adaptive capacity, low total attributable fraction;
Weaknesses (W) reflect current vulnerabilities — high population sensitivity, high heat- and cold-attributable fractions;
Opportunities (O) identify windows for intervention — municipalities with a relatively low vulnerability index and moderate current exposure have room to advance prevention before reaching critical levels;
Threats (T) signal emerging or already present risks — high climate exposure, high relative heat risk (DLNM), high attributable number.

Data used

Object	Source	Content
`caso_espacial.rds`	Tutorial T5	sf geometries of the RMSP municipalities
`caso_socio.rds`	Tutorial T6	Socioeconomic indicators per municipality (HDI, % elderly, etc.)
`mod_dlnm_fit.rds`	Module M1	`climasus_dlnm` object fitted for the RMSP
`mod_af.rds`	Module M2	`climasus_af` object with total/heat/cold AF%

Whenever any of these four files is missing (execution outside the full chain), the pre-rendering script synthesizes realistic data with a warning — the Rmd renders normally, displaying illustrative results.

Complete pipeline

library(climasus4r)
library(dplyr)

Load chained inputs

# Objeto climasus_af do Módulo M2
mod_af <- if (file.exists("vignettes-pt/dados/mod_af.rds")) {
  readRDS("vignettes-pt/dados/mod_af.rds")
} else {
  NULL  # fallback: SWOT rodará sem AF
}

# Objeto climasus_dlnm do Módulo M1
mod_dlnm <- if (file.exists("vignettes-pt/dados/mod_dlnm_fit.rds")) {
  readRDS("vignettes-pt/dados/mod_dlnm_fit.rds")
} else {
  NULL  # fallback: sem extração DLNM automática
}

# Indicadores socioeconômicos
caso_socio <- if (file.exists("vignettes-pt/dados/caso_socio.rds")) {
  readRDS("vignettes-pt/dados/caso_socio.rds")
} else {
  # Mínimo realista para o script funcionar sem a cadeia completa
  data.frame(
    city          = c("Sao Paulo", "Guarulhos", "Santo Andre", "Osasco", "Maua"),
    pct_idosos    = c(16.2, 12.4, 14.8, 13.1, 11.9),
    pct_doenca_cr = c(28.1, 24.3, 26.7, 22.8, 20.5),
    hdi           = c(0.845, 0.763, 0.815, 0.776, 0.742),
    pct_agua      = c(97.2, 88.4, 94.1, 90.3, 85.7),
    densidade_ubs = c(6.8, 3.2, 5.1, 4.0, 2.9)
  )
}

Build the three IPCC pillars

Each pillar is supplied as a data.frame with a municipal identifier column (city_col = "city") and numeric indicator columns. Indicators missing for a given municipality are filled with NA and excluded from the pillar’s weighted average, without forcing complete exclusion.

# ----- Pilar E: Exposição Climática ------------------------------------------
# Fonte: sus_climate_aggregate() (temperatura maxima media, dias de calor P90,
#        contagem de ondas de calor por ano)
exposure_df <- data.frame(
  city           = caso_socio$city,
  tmax_media     = c(29.8, 31.2, 28.9, 30.1, 31.8),  # grau Celsius
  dias_calor_p90 = c(52,   68,   45,   58,   72),     # dias/ano
  ondas_calor    = c(4,    8,    3,    6,    9)        # eventos/ano
)

# ----- Pilar S: Sensibilidade Populacional -----------------------------------
# Fonte: caso_socio.rds + extração automática de climasus_dlnm (argumento fits)
sensitivity_df <- caso_socio[, c("city", "pct_idosos", "pct_doenca_cr")]

# ----- Pilar AC: Capacidade Adaptativa ---------------------------------------
# Fonte: caso_socio.rds (HDI, cobertura hídrica, densidade UBS)
adaptive_capacity_df <- caso_socio[, c("city", "hdi", "pct_agua", "densidade_ubs")]

Fit the Composite Vulnerability Index

sus_mod_vulnerability_index() accepts a fits argument (a named list of climasus_dlnm objects) for automatic extraction of the metrics dlnm_hot_rr, dlnm_cold_rr, and dlnm_sensitivity_index, which are automatically merged into the sensitivity pillar.

# Opcional: lista nomeada de climasus_dlnm (um por cidade)
# fits_dlnm <- list(sao_paulo = mod_dlnm)  # ajuste por municipio

vi_fit <- sus_mod_vulnerability_index(
  exposure_df          = exposure_df,
  sensitivity_df       = sensitivity_df,
  adaptive_capacity_df = adaptive_capacity_df,
  fits                 = NULL,                # ou: fits_dlnm (lista climasus_dlnm)
  city_col             = "city",
  normalize            = "minmax",            # "minmax" ou "zscore"
  weights              = c(tmax_media = 2,    # dobrar peso da Tmax
                           pct_idosos = 2),   # dobrar peso da proporção de idosos
  pillar_weights       = c(exposure         = 1.5,
                           sensitivity      = 1.0,
                           adaptive_capacity = 1.0),
  hot_percentile       = 0.99,
  cold_percentile      = 0.01,
  lang                 = "pt",
  verbose              = TRUE
)

# Inspecionar o resultado
print(vi_fit)

The climasus_vi object contains:

Element	Type	Content
`$vi_table`	tibble	CVI, pillar scores, rank, and percentile per city
`$pillar_scores`	tibble (long)	Raw scores for the three pillars per city
`$normalized_indicators`	tibble	Normalized indicators (prefixes `exp_`, `sen_`, `ac_`)
`$raw_indicators`	tibble	Original indicators without normalization
`$weights`	list	Intra-pillar and inter-pillar weights effectively used
`$gini_coefficient`	numeric	Gini of the CVI (0 = perfect equality)
`$meta`	list	Call parameters, counts, timing

Fit the SWOT Analysis

sus_mod_swot() aggregates climasus objects from multiple modules into normalized (0–100) and categorized SWOT quadrants. The argument breaks = c(25, 50, 75) produces four categories (Low / Medium / High / Very High) — the default c(33, 66) produces three categories.

swot_fit <- sus_mod_swot(
  vulnerability = vi_fit,
  af            = mod_af,    # climasus_af do Módulo M2 (ou NULL)
  burden        = NULL,      # climasus_burden do Módulo M2 (opcional)
  dlnm          = mod_dlnm,  # climasus_dlnm do Módulo M1 (broadcast)
  sensitivity   = NULL,      # climasus_sensitivity do Módulo M7 (opcional)
  score_type    = "both",    # "numeric", "categorical" ou "both"
  breaks        = c(25, 50, 75),
  labels        = c("Baixo", "Médio", "Alto", "Muito Alto"),  # 4 cats para 3 breaks
  city_col      = "city",
  lang          = "pt",
  verbose       = TRUE
)

# Visualizar tabela de escores
swot_fit$scores

Function characteristics

`sus_mod_vulnerability_index()` — IPCC Vulnerability Index

Full signature:

sus_mod_vulnerability_index(
  exposure_df          = NULL,
  sensitivity_df       = NULL,
  adaptive_capacity_df = NULL,
  fits                 = NULL,
  city_col             = "city",
  normalize            = "minmax",
  weights              = NULL,
  pillar_weights       = c(exposure = 1, sensitivity = 1, adaptive_capacity = 1),
  hot_percentile       = 0.99,
  cold_percentile      = 0.01,
  lang                 = "pt",
  verbose              = TRUE
)

Key parameters:

Argument	Type	Description
`exposure_df`	`data.frame`	Indicators for the Exposure pillar (higher = more exposed)
`sensitivity_df`	`data.frame`	Indicators for the Sensitivity pillar (higher = more sensitive)
`adaptive_capacity_df`	`data.frame`	Indicators for the Adaptive Capacity pillar (higher = more adapted)
`fits`	named list of `climasus_dlnm`	Automatic extraction of heat/cold RR and sensitivity index
`city_col`	`character`	Name of the municipal identifier column (default `"city"`)
`normalize`	`"minmax"` or `"zscore"`	Indicator normalization method
`weights`	named vector	Intra-pillar weights by column name; unlisted columns get weight 1
`pillar_weights`	named vector `c(exposure, sensitivity, adaptive_capacity)`	Inter-pillar weights; missing pillars automatically get weight 0
`hot_percentile` / `cold_percentile`	numeric	Quantiles for RR extraction from the DLNM fits

Return — climasus_vi object:

$vi_table: tibble ordered by vi_rank (rank 1 = most vulnerable), with columns vi_score ∈ [0, 1], exposure_score, sensitivity_score, adaptive_capacity_score, vi_rank, vi_percentile.
$pillar_scores: long tibble with city, pillar, score (raw score before the AC inversion).
$normalized_indicators / $raw_indicators: wide tibbles with prefixes exp_, sen_, ac_.
$weights: list with the intra-pillar and inter-pillar weights effectively used.
$gini_coefficient: Gini of the CVI distribution.
$meta: list of parameters, counts, and timestamp.

Available S3 methods: print(), summary(), tidy() (returns a flat tibble), coef() (returns a named vector city → vi_score).

`sus_mod_plot_vulnerability()` — CVI Visualizations

Full signature:

sus_mod_plot_vulnerability(
  x,
  type        = c("ranking", "pillars", "lorenz"),
  output_type = c("plot", "table", "all"),
  interactive = FALSE,
  base_size   = 12L,
  save_plot   = NULL,
  lang        = c("pt", "en", "es"),
  verbose     = FALSE
)

Plot types (type=):

`type`	Description	Data returned in `"table"`
`"ranking"`	Horizontal lollipop: CVI per city, most vulnerable at the top	`$vi_table`
`"pillars"`	Horizontal stacked bars: contribution of E / S / AC per city	`$pillar_scores`
`"lorenz"`	Lorenz curve of CVI inequality with the Gini annotated	`$vi_table`

Return: ggplot (output_type = "plot"), tibble ("table"), or a named list $plot, $table, $data ("all"). With interactive = TRUE, returns a plotly widget (requires the plotly package).

`sus_mod_swot()` — Climate-Health SWOT Synthesis

Full signature:

sus_mod_swot(
  vulnerability = NULL,
  af            = NULL,
  burden        = NULL,
  dlnm          = NULL,
  sensitivity   = NULL,
  score_type    = c("both", "numeric", "categorical"),
  breaks        = c(33, 66),
  labels        = NULL,
  city_col      = "city",
  lang          = c("pt", "en", "es"),
  verbose       = TRUE
)

Indicators per quadrant:

Quadrant	Source	Extracted indicators
S Strengths	`climasus_vi`	`adaptive_capacity_score` (high capacity = strength)
S Strengths	`climasus_af`	100 − total AF% (low burden = strength)
S Strengths	`climasus_burden`	Inverted burden rank
W Weaknesses	`climasus_vi`	`sensitivity_score` (high sensitivity = weakness)
W Weaknesses	`climasus_af`	Heat AF%; cold AF%
W Weaknesses	`climasus_sensitivity`	RR inequality across strata
O Opportunities	`climasus_vi`	100 − CVI percentile (low percentile = window); current exposure inverted
T Threats	`climasus_vi`	`exposure_score`; `vi_score`
T Threats	`climasus_dlnm`	Heat RR at P95 (scaled 0–100)
T Threats	`climasus_burden`	Normalized attributable number
T Threats	`climasus_sensitivity`	Maximum RR across strata

Return — climasus_swot object:

$scores: tibble with one row per entity; columns entity, S_score, W_score, O_score, T_score (0–100) and, when score_type %in% c("categorical","both"), S_cat, W_cat, O_cat, T_cat; counts n_S, n_W, n_O, n_T.
$indicators: long tibble with one row per (entity × quadrant × indicator): entity, quadrant, ind_code, indicator, raw_value, norm_score, direction ("positive" or "negative").
$meta: list with n_entities, n_indicators, inputs_used, score_type, breaks, labels, lang, call_time.

S3 methods: print(), summary(), tidy() (returns $scores).

`sus_mod_plot_swot()` — SWOT Visualizations

Full signature:

sus_mod_plot_swot(
  x,
  type        = c("radar", "matrix", "bar"),
  output_type = c("plot", "table", "all"),
  interactive = FALSE,
  entities    = NULL,
  top_n       = 3L,
  base_size   = 12L,
  save_plot   = NULL,
  lang        = c("pt", "en", "es"),
  verbose     = FALSE
)

Plot types (type=):

`type`	Description	Recommended use
`"radar"`	Spider/radar chart of the four SWOT scores (0–100) per entity, with overlapping polygons for multi-city comparison	Comparing profiles across municipalities
`"matrix"`	Classic 2×2 chart: numeric score, category, and bullet points of the leading indicators per quadrant. Shows only one entity (the first in the `entities` selection)	Risk communication for a single municipality
`"bar"`	Grouped horizontal bars of the four SWOT scores per entity	Quick comparison across all municipalities

Argument top_n (only for type = "matrix"): maximum number of indicator bullets displayed per quadrant (default 3).

Return: ggplot or plotly (output_type = "plot"), tibble ("table"), or a list $plot, $table ("all").

Reference formulas

Composite Vulnerability Index (IPCC)

The central formula combines the three pillars with the Adaptive Capacity inversion:

$VI_i = \frac{w_E \cdot E_i + w_S \cdot S_i + w_{AC} \cdot (1 - AC_i)}{w_E + w_S + w_{AC}}$

where each pillar $P_i$ is the weighted average of the normalized indicators:

$P_i = \frac{\sum_j \omega_j \tilde{x}_{ij}}{\sum_j \omega_j}$

Missing indicators are excluded and the remaining weights are renormalized, partially preserving municipalities with incomplete data.

Min-max normalization

$\tilde{x}_{ij} = \frac{x_{ij} - \min_j(x_{ij})}{\max_j(x_{ij}) - \min_j(x_{ij})}$

When $\max_j - \min_j \approx 0$ (constant variable), the value is fixed at 0.5 (implemented in .vi_minmax_col()).

Gini Coefficient

Given the vector $\mathbf{v}$ of $VI_i$ scores sorted in ascending order:

$G = \frac{2 \sum_{i=1}^{n} i \cdot v_i}{n \sum_{i=1}^{n} v_i} - \frac{n+1}{n}$

The Gini is reported in $gini_coefficient and annotated on the type = "ranking" and type = "lorenz" plots.

Illustrated results

Vulnerability ranking

The lollipop chart orders municipalities from most (top) to least vulnerable, with the CVI score on the horizontal axis. The Gini Coefficient in the subtitle quantifies the territorial inequality of vulnerability.

Figure 1. Climate-health vulnerability ranking of the RMSP municipalities. Min-max scores on the x-axis; the municipality at the top = most vulnerable (highest CVI). Gini indicates the concentration of risk across municipalities.

Decomposition of the IPCC pillars

The stacked bars allow identification of which pillar dominates the vulnerability of each municipality: Exposure (red), Sensitivity (blue), or Adaptive Capacity (green). Municipalities with balanced pillars indicate multifactorial risk.

Figure 2. Decomposition of the three IPCC pillar scores by municipality, ordered by total CVI (most vulnerable at the top).

Lorenz curve of vulnerability

The Lorenz curve accumulates the proportion of total CVI (y-axis) as a function of the proportion of municipalities (x-axis), from least to most vulnerable. The farther from the diagonal (perfect equality), the greater the concentration of risk in a few municipalities.

Figure 3. Lorenz curve of the Composite Vulnerability Index in the RMSP. The shaded area between the curve and the diagonal of perfect equality (dashed) is proportional to the Gini Coefficient annotated on the plot.

Vulnerability score table

Table 1. Vulnerability scores by municipality (top 10, decreasing order of vulnerability).
city	vi_score	exposure_score	sensitivity_score	adaptive_capacity_score	vi_rank	vi_percentile
350950	0.8069	0.7717	0.7500	0.0836	1	100.0
351880	0.5076	0.4375	0.5713	0.4510	2	83.3
rmsp	0.5000	NA	0.5000	NA	3	66.7
355030	0.4317	0.4033	0.5352	0.6291	4	50.0
351380	0.3617	0.2576	0.5165	0.6369	5	33.3
354870	0.1561	0.1069	0.3283	0.9421	6	16.7

SWOT profile — Radar

The radar chart shows the four SWOT scores (0–100) for all municipalities simultaneously. Larger polygons (especially in W and T) signal greater fragility; wide polygons in S and O indicate protective assets and intervention opportunities.

Figure 4. SWOT profile by municipality in the RMSP. Axes: Strengths (top), Opportunities (right), Threats (bottom), Weaknesses (left). Scale 0–100.

SWOT 2×2 chart — Most vulnerable municipality

The classic 2×2 chart displays, for the municipality with the highest CVI, the numeric score and category of each quadrant, together with the leading indicators (with normalized score) that contributed the most to each quadrant.

Figure 5. SWOT 2×2 chart for the most vulnerable municipality in the RMSP. An ideal format for risk communication to municipal managers.

SWOT scores — Comparative bars

Figure 6. Grouped SWOT scores by municipality. Allows simultaneous comparison of Strengths, Weaknesses, Opportunities, and Threats across all municipalities in the RMSP.

SWOT score table

Table 2. SWOT scores and categories by municipality in the RMSP.
entity	S_score	n_S	W_score	n_W	O_score	n_O	T_score	n_T	S_cat	W_cat	O_cat	T_cat
350950	40.44680	2	6.643333	3	49.61415	2	6.898991	3	Cat2	Cat1	Cat2	Cat1
351880	40.63050	2	6.583767	3	58.13125	2	6.687824	3	Cat2	Cat1	Cat3	Cat1
rmsp	80.81000	2	6.560000	3	33.30000	2	9.809187	3	Cat4	Cat1	Cat2	Cat1
355030	40.71955	2	6.571733	3	74.79835	2	6.651124	3	Cat2	Cat1	Cat3	Cat1
351380	40.72345	2	6.565500	3	83.22120	2	6.579224	3	Cat2	Cat1	Cat4	Cat1
354870	40.87605	2	6.502767	3	91.59655	2	6.460458	3	Cat2	Cat1	Cat4	Cat1

References

Andrews, K. (1971). The Concept of Corporate Strategy. Dow Jones-Irwin.

Benmarhnia, T., Deguen, S., Kaufman, J. S., & Smargiassi, A. (2015). Vulnerability to heat-related mortality: a systematic review, meta-analysis, and meta-regression analysis. Epidemiology, 26(6), 781–793. https://doi.org/10.1097/EDE.0000000000000375

Confalonieri, U., Marinho, D., Dias, H., & Malheiros, T. (2009). O índice de vulnerabilidade aos riscos climáticos do Brasil. Parcerias Estratégicas, 27, 161–172.

Cutter, S. L., Boruff, B. J., & Shirley, W. L. (2003). Social vulnerability to environmental hazards. Social Science & Medicine, 56(11), 2301–2322. https://doi.org/10.1016/S0277-9536(02)00605-X

Gasparrini, A., & Armstrong, B. (2013). Attributable risk from distributed lag models. BMC Medical Research Methodology, 13, 3. https://doi.org/10.1186/1471-2288-13-3

Gasparrini, A., et al. (2017). Projections of temperature-related excess mortality under climate change scenarios. The Lancet Planetary Health, 1(9), e360–e367. https://doi.org/10.1016/S2542-5196(17)30156-0

IPCC. (2014). Fifth Assessment Report — Working Group II: Impacts, Adaptation and Vulnerability. Part A: Global and Sectoral Aspects. Chapter 19: Emergent Risks and Key Vulnerabilities. Cambridge University Press. https://doi.org/10.1017/CBO9781107415379

IPCC. (2022). Sixth Assessment Report — Working Group II: Impacts, Adaptation and Vulnerability. Chapter 16: Key Risks across Sectors and Regions. Cambridge University Press. https://doi.org/10.1017/9781009325844

Marengo, J. A., et al. (2022). Vulnerability to climate change in Brazil: current status and future scenarios. International Journal of Climate Change Strategies and Management, 14(2), 133–149.

Turner, B. L., et al. (2003). A framework for vulnerability analysis in sustainability science. PNAS, 100(14), 8074–8079. https://doi.org/10.1073/pnas.1231335100

End of the Modeling section. This is the final module in the Modeling series of climasus4r. The results produced here (mod_vi.rds) can be loaded into final reports and risk communication for managers and decision-makers.

Previous module	Modeling 07: Spatial and Spatio-Temporal Bayes
Tutorial hub	Tutorial Index

How to reproduce the figures. Run once, from the package root: Rscript scripts/modelagem_08_vulnerabilidade_swot.R. The script saves the figures in vignettes-pt/figuras/ (prefixes modvi_ and modswot_), the tables in vignettes-pt/dados/, and the chained vulnerability object (mod_vi.rds).

climasus4r Team

17/10/2026

Objective

Scientific background

Climate vulnerability according to the IPCC

Min-max and z-score normalization

Territorial inequality and the Lorenz Curve

SWOT analysis oriented toward health-climate policy

Data used

Complete pipeline

Load chained inputs

Build the three IPCC pillars

Fit the Composite Vulnerability Index

Fit the SWOT Analysis

Function characteristics

`sus_mod_vulnerability_index()` — IPCC Vulnerability Index

`sus_mod_plot_vulnerability()` — CVI Visualizations

`sus_mod_swot()` — Climate-Health SWOT Synthesis

`sus_mod_plot_swot()` — SWOT Visualizations

Reference formulas

Composite Vulnerability Index (IPCC)

Min-max normalization

Gini Coefficient

Illustrated results

Vulnerability ranking

Decomposition of the IPCC pillars

Lorenz curve of vulnerability

Vulnerability score table

SWOT profile — Radar

SWOT 2×2 chart — Most vulnerable municipality

SWOT scores — Comparative bars

SWOT score table

References

Navigation

Modeling 08: Vulnerability Index and SWOT Synthesis

climasus4r Team

17/10/2026

Objective

Scientific background

Climate vulnerability according to the IPCC

Min-max and z-score normalization

Territorial inequality and the Lorenz Curve

SWOT analysis oriented toward health-climate policy

Data used

Complete pipeline

Load chained inputs

Build the three IPCC pillars

Fit the Composite Vulnerability Index

Fit the SWOT Analysis

Function characteristics

sus_mod_vulnerability_index() — IPCC Vulnerability Index

sus_mod_plot_vulnerability() — CVI Visualizations

sus_mod_swot() — Climate-Health SWOT Synthesis

sus_mod_plot_swot() — SWOT Visualizations

Reference formulas

Composite Vulnerability Index (IPCC)

Min-max normalization

Gini Coefficient

Illustrated results

Vulnerability ranking

Decomposition of the IPCC pillars

Lorenz curve of vulnerability

Vulnerability score table

SWOT profile — Radar

SWOT 2×2 chart — Most vulnerable municipality

SWOT scores — Comparative bars

SWOT score table

References

Navigation

`sus_mod_vulnerability_index()` — IPCC Vulnerability Index

`sus_mod_plot_vulnerability()` — CVI Visualizations

`sus_mod_swot()` — Climate-Health SWOT Synthesis

`sus_mod_plot_swot()` — SWOT Visualizations