## wpiea2026083-source-pdf

## Source details

**Canonical URL:** [wpiea2026083-source-pdf](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026083-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026083-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026083-source-pdf.pdf.json)

---

### Key findings on adaptation needs
- Pacific Island Countries (PICs) face acute and rising climate adaptation needs owing to exposure to sea‑level rise, natural hazards, and structural vulnerabilities associated with small, remote island economies.
- For atoll countries such as Kiribati, Tuvalu, and the Marshall Islands, projected sea‑level rise of up to 0.5 meters by 2070–2110 could submerge large portions of urban areas.
- Physical risks—particularly from coastal inundation and sea‑level rise—account for the bulk of identified needs, with the highest burdens falling on atoll nations.
- When expressed relative to GDP, adaptation needs are exceptionally high in most PICs, reflecting both the magnitude of required investments and the region’s relatively small economic base.
- Existing assessments vary widely in scope, methodology, sectoral coverage, and time horizon, creating challenges for aggregation and comparison.

### Methodological approach
- Unified framework combining:
  - A metadata analysis that assembles information from existing country reports and standardizes it into comparable annual measures of adaptation needs.
  - A complementary machine‑learning approach to generate synthetic estimates for data‑deficient countries and to assess internal coherence across metadata‑based estimates.
- Metadata framework converts estimates into annual percent‑of‑GDP and 2024 USD values and aligns time horizons for regional analysis.
- Reported estimates are conditional on protection targets, time horizons, climate scenarios, and financing conditions.
- Table 2.1 summary statistics (selected):
  - Fiji: No. of Reports 3; No. of Sectors 14; End of Timeframe 2060
  - Kiribati: No. of Reports 4; No. of Sectors 20; End of Timeframe 2150
  - Marshall Islands: No. of Reports 3; No. of Sectors 7; End of Timeframe 2150
  - Micronesia: No. of Reports 4; No. of Sectors 13; End of Timeframe 2060
  - Nauru: No. of Reports 1; No. of Sectors 1; End of Timeframe 2035 (note: for Nauru, there are a total of 9 sectors, but only 1 extends until 2035, while the others end in 2023)
  - Palau: No. of Reports 3; No. of Sectors 9; End of Timeframe 2060
  - Papua New Guinea: No. of Reports 2; No. of Sectors 5; End of Timeframe 2030
  - Samoa: No. of Reports 3; No. of Sectors 12; End of Timeframe 2060
  - Solomon Islands: No. of Reports 4; No. of Sectors 10; End of Timeframe 2060
  - Tonga: No. of Reports 4; No. of Sectors 15; End of Timeframe 2060
  - Tuvalu: No. of Reports 4; No. of Sectors 13; End of Timeframe 2150
  - Vanuatu: No. of Reports 3; No. of Sectors 28; End of Timeframe 2060

### Adaptation finance flows and delivery
- Regional and global headline figures:
  - Global climate flows reached USD 136 billion in 2022.
  - Out of that total, USD 65 billion were commitments with adaptation as a principal or significant objective.
  - Paris agreement commitment: US$ 100 billion annual goal.
- Adaptation finance in PICs:
  - In 2023, adaptation was the objective in 90 percent of all climate finance commitments to PICs.
  - PICs received about 2 percent of global adaptation finance commitments over the 2019-2023 period.
  - Papua New Guinea accounted for a third of all adaptation finance commitments to PICs and represents 73 percent of PICs’ GDP and 79 percent of PICs’ population.
- Temporal pattern:
  - Adaptation finance commitments accelerated over the past decade: more commitments over the 2019-2023 period than over the eight years prior.
  - 2022 marked a milestone as global climate finance targets were achieved and surpassed.
- Composition by instrument (2011-2023):
  - Grants represented 84 percent of total adaptation finance commitments.
  - Half of PICs relied solely or almost solely on grants.
  - The four countries that relied the most on loans are the four larger economies in the region.
- Composition by source (2011-2023):
  - 60 percent of adaptation finance commitments to the PIC region came from bilateral donors.
  - 40 percent came from multilateral donors.
  - Multilateral composition (share of commitments):
    - World Bank: 46%
    - ADB: 19%
    - Green Climate Fund: 18%
    - Other Climate Funds: 15%
    - Other MDBs: 2%
  - Finance from private foundations or nonprofit organizations represented a very limited share.
- Country-level commitments (2023 USD billion; estimates exclude regional projects):
  - Values reported in the figure: 2.15, 0.99, 0.84, 0.75, 0.57, 0.41, 0.36, 0.30, 0.25, 0.23, 0.14, 0.07
- Disbursements versus commitments:
  - Disbursements correspond to actual payments; commitments are recorded at signature and reflect full project cost.
  - Rapid increases in commitments typically translate into higher disbursements only after a few years.
  - Disbursement ratios vary significantly across financing channels:
    - Generally higher for bilateral partners.
    - Generally lower for multilateral institutions and climate funds.
  - Variation reflects differences in project design, approval processes, and implementation capacity.
- Data caveat:
  - Commitment data are self‑reported and rely on tracking methodologies with limitations; commitment data likely to be inflated.

### Determinants of finance mobilization
- Vulnerability to climate change remains a strong determinant of resource allocation.
- Institutional capacity and governance readiness appear to influence financing outcomes.
- Weak capacity continues to constrain implementation in several countries.
- External factors—most notably fiscal pressures in provider countries—may affect the availability of future climate finance.

### Adaptation financing gaps and implications
- Harmonized, region‑wide estimates show large and uneven adaptation needs across PICs.
- Current financing commitments and disbursements are insufficient to meet projected adaptation needs, creating substantial adaptation‑financing gaps.
- Differences in source material, potential overlaps in reported cost components, and extended time horizons in some assessments introduce uncertainty and warrant caution in cross‑country interpretation.

### Policy implications and priorities
- Urgent expansion and acceleration of adaptation finance for Pacific Island Countries are required.
- Key priorities include:
  - Enhancing institutional readiness.
  - Strengthening public financial management for climate‑related investments.
  - Improving the effectiveness and accessibility of financing mechanisms, including emerging regional platforms such as the Pacific Resilience Facility.
  - Addressing implementation capacity constraints to improve disbursement performance and close financing gaps.

### Contributions of the paper
- Develops a unified framework to organize and standardize climate adaptation needs drawn from diverse country reports, enabling consistent cross‑country comparison.
- Complements the metadata analysis with a machine‑learning approach to address gaps for data‑deficient countries and provide interpretive perspective.
- Brings together harmonized adaptation‑needs estimates with detailed evidence on climate finance commitments and disbursements to PICs, offering a new assessment of adaptation financing gaps and factors associated with effective access to climate finance.

### Data sources and literature coverage
- Primary cited analyses used for adaptation financing needs:
  - IMF departmental paper on Fiscal Policy to Address Climate Change in Asia and the Pacific (Dabla-Norris et al. (2021)) — covers most PIC countries except Nauru.
  - World Bank Country Climate and Development Report (CCDR) for the Pacific Atoll Countries (World Bank (2024)) — covers Kiribati, Tuvalu, and the Marshall Islands.
  - World Bank Climate Change and Disaster Management (CCDM) report (2016) — addresses most PICs, except Papua New Guinea and Nauru.
- UNEP’s Adaptation Gap Report (2023) presents aggregated estimates but is not cited directly because it is based on the above country-specific reports.
- Annex I lists country-specific reports used as sources for the metadata analysis.

### Matrix framework and data processing steps
- Purpose: estimate adaptation needs for each country using a standardized matrix to enable regional summation.
- Steps:
  - Step 1: Search reports to determine which sectors are assessed for financing needs; record sector names and exact amounts in original units (percentage of GDP or U.S. dollar values); enter each sector as a separate row with corresponding values.
  - Step 2: Record time span for each estimate and standardize estimates as annual percentage of GDP in one column.
  - Step 3: Convert standardized annual cost into 2024 USD millions by:
    - Applying actual U.S. inflation rate to rebase estimates expressed in past years.
    - Applying a 2 percent inflation assumption (U.S. long-run inflation target) to extrapolate estimates expressed in future years.
- Note: Conversion abstracts from uncertainty in future inflation, exchange rates, and financing conditions and is intended to provide order-of-magnitude estimates rather than precise projections.

### Sector classification and aggregation
- Original reports yield 56 sector categories; the framework preserves original terminology (e.g., "education" and "social spending – education" kept separate).
- Categories pooled into two groups corresponding to the risks they are adapted to.
- For concise assessment, detailed sectors summarized into:
  - Primary sectors: capture largest adaptation needs from direct exposure of physical capital to natural disasters and rising sea levels (e.g., coastal protection infrastructure, land reclamation).
  - Secondary sectors: all other sectors (examples include Agriculture, Tourism Resilience, Biosecurity, Ocean and Marine Protection, Wetland Conservation, Capacity Building, Climate Policy, Education, Health Services, Social Protection, Water and Sanitation).
- Rationale: largest cost estimates in key source reports are concentrated in primary sectors; reclassifying small secondary components would not materially affect aggregate estimates.

### Caveats and methodological limitations
- Potential overlap or double counting of financing needs estimates across sources.
- Possible omission of relevant literature.
- Sources differ in scope and methodology (from incremental resilience upgrades to comprehensive adaptation), so large GDP shares do not imply full climate insulation.
- The cost of reallocating the entire population to a newly established site in another country lies beyond the scope of existing analysis and is not captured.
- Some sectors may be misclassified; however, main cost drivers are in primary sectors.

### Findings: Annual financing needs (metadata analysis)
- Regional total (Pacific Island Countries, PICs):
  - Annual financing need around USD 3.3 billion (at 2024 prices).
  - Physical risks account for 56 percent of needs.
  - Other risks account for 44 percent of needs.
- Country-level annual needs (2024 USD million):
  - Papua New Guinea: USD 868 million.
  - Fiji: USD 857 million.
  - Samoa, Micronesia, Tonga, the Marshall Islands, Kiribati: needs above 100 million and below 200 million.
  - Nauru: only one study available for needs beyond 2023 and covers only other risks.
- Regional needs as percent of GDP:
  - Average annual needs: 22.5 percent of GDP.
  - Physical risks: 13.1 percent of GDP.
  - Other risks: 9.4 percent of GDP.
  - Footnote: This amounts to 24.5 percent of GDP when using the random‑forest–estimated needs for Nauru, as reported in Section 2.2.3.
- Sectoral patterns by country:
  - Physical risks dominate in Tuvalu (97 percent), the Marshall Islands (85 percent), and Kiribati, Solomon Islands, Palau, and Papua New Guinea (all above 60 percent).
  - Other risks dominate in Micronesia, Tonga, and Samoa.

### Time horizons covered by assessments
- Time horizons vary significantly across assessments: range from 10 to 125 years.
- Physical risks:
  - Often projected over much longer periods; for the three atoll countries (Kiribati, the Marshall Islands, Tuvalu) assessments consider a two-meter sea-level rise beyond 2150, yielding a 125-year time span.
  - In non-atoll countries, World Bank CCDM typically uses a 40-year horizon, divided into 2020–2040 and 2040–2060.
- Other risks:
  - Generally assessed over the medium term—up to 2050; many other assessments adopt approximately a 10-year horizon.
- Examples of assessment time spans:
  - IMF Working Paper on Fiscal Policies to Address Climate Change in Asia and the Pacific: 2021–2030.
  - Kiribati Joint Implementation Plan for Climate Change and Disaster Risk Management: 2013–2023.
  - IMF Technical Assistance Report on Climate Change Policy in Tonga: 2018–2028.
  - Fiji’s Climate Vulnerability Assessment: 2017–2027.
- Data gaps:
  - Nauru: most identified financing needs end in 2023, with one analysis on water and sanitation extending to 2035.
  - Papua New Guinea: no financing needs assessment extending beyond 2030.

### Machine learning approach for data‑deficient countries
- Purpose:
  - Address heterogeneity in coverage, timespan, and methodology across literature.
  - Generate synthetic estimates for data‑deficient countries (e.g., Nauru) based on relationships between adaptation needs and country characteristics.
- Model choice and rationale:
  - Random forest model selected for relative robustness to small sample size.
  - Training data: 11 countries (Nauru excluded from training).
  - Due to small sample, no train/test split; visual checks used for relationships between dependent and important independent variables.
  - To address uncertainty, results averaged from 200 random forests (Bayesian spirit).
  - Each forest:
    - Comprises 1000 trees.
    - Each tree trained on 4 independent variables randomly selected from a pool of 24 independent variables.
    - Unlimited tree depth.

### Random forest model inputs and dependent variable
- Dependent variable:
  - Annualized adaptation needs (percentage of GDP, annualized).
  - Table 2.3 summary statistics (as presented):
    - Depedent Var. Adaptaion needs (% of GDP, annualized) 11 24.2 12.5
- Explanatory variables (24 total) span economic indicators, geographic attributes, and climate‑related vulnerability metrics. Examples and summary statistics (as presented) include:
  - GDP per capita (US$): N 11, Mean 12.33, Std. Dev. 3893.8
  - Nominal GDP (US$ mil.): N 11, Mean 3207.47, Std. Dev. 47307.0
  - Public debt (% of GDP): N 11, Mean 36.9, Std. Dev. 26.6
  - Population (thousands): N 11, Mean 35436, Std. Dev. 37
  - Tourism, dummy: N 11, Mean 0.45, Std. Dev. 0.52
  - Distance to economic centers: N 11, Mean 168.7, Std. Dev. 12.3
  - Distance to nearest (non-SIDs) neighbor: N 11, Mean 57.2, Std. Dev. 28.5
  - Distance to trading partners: N 11, Mean 50.5, Std. Dev. 7.8
  - Liner Shipping Connectivity: N 11, Mean 24.5, Std. Dev. 13.8
  - Freight costs of imports (% of good imports): N 11, Mean 14.2, Std. Dev. 5.7
  - Atoll, dummy: N 11, Mean 0.27, Std. Dev. 0.47
  - Disaster Severity (% of population): N 11, Mean 5.2, Std. Dev. 18.2
  - Disasters Frequency, Flood: N 11, Mean 0.14, Std. Dev. 0.20
  - Disasters Frequency, Storm: N 11, Mean 0.27, Std. Dev. 0.24
  - Disasters Frequency, Total: N 11, Mean 0.50, Std. Dev. 0.42
  - Expected Severity (% of population): N 11, Mean 4.55, Std. Dev. 3.14
  - Climate risk: N 11, Mean 3.7, Std. Dev. 0.70
  - Climate risk, Hazard & Exposure: N 11, Mean 2.5, Std. Dev. 0.76
  - Climate risk, Vulnerability: N 11, Mean 4.1, Std. Dev. 0.60
  - Lack of Coping capacity: N 11, Mean 5.0, Std. Dev. 1.30
  - Readiness: N 11, Mean 0.43, Std. Dev. 0.09
  - Economic Readiness: N 10, Mean 0.44, Std. Dev. 0.14
  - Vunerability: N 9, Mean 0.55, Std. Dev. 0.06
  - Exposure: N 11, Mean 0.54, Std. Dev. 0.06
- Note: Table AII.2 in Annex II provides sources and definitions of variables.

### Machine learning training results and interpretive findings
- Overall:
  - Adaptation needs are significantly correlated with most explanatory variables.
  - Figure 2.6 identifies relative importance of explanatory variables; eight variables stand out, while others remain relatively important.
- Key observed relationships (visual assessment):
  - Nominal GDP: strong negative relationship with adaptation needs — smaller economies exhibit higher required adaptation investments per GDP.
  - Public debt: negative correlation with adaptation needs — low public debt may reflect limited borrowing capacity and higher unmet needs.
  - Geographic remoteness and higher shipping costs: positive association with higher adaptation investment costs.
  - Climate-related vulnerabilities (e.g., exposure to natural disasters): positive relationship with adaptation needs.
  - Atoll countries tend to face substantial adaptation needs, reflecting fragility driven by small size, remoteness, high shipping costs, and greater exposure to disasters.
- Interpretation:
  - Expected associations from the random forest visual checks support the plausibility of metadata-derived estimates despite heterogeneity in source coverage, time horizons, and methodologies.
- Uncertainty and limitations:
  - The prominence of many variables may partly reflect estimation uncertainty given small sample size.
  - For Nauru, large uncertainty in metadata estimates led to omission of the metadata estimate; the machine learning approach generates a synthetic estimate for data-deficient cases.

### Prediction of Missing Annual Adaptation Needs (Imputed 2024 annualized percent of GDP)
- Random-forest imputed adaptation needs (In percent of GDP, annualized, estimated annual requirement for 2024):
  - 35.7
  - 34.3
  - 32.7
  - 30.1
  - 28.3
  - 27.9
  - 23.6
  - 23.0
  - 21.3
  - 17.1
  - 14.7
  - 11.5
- Nauru (NRU) is predicted to require the fifth-largest adaptation investment as a percentage of GDP among the 12 PICs (Figure 2.8, orange).
- The random forest projection preserves the rankings from the metadata analysis, with mid-ranked countries showing minimal discrepancies between methods.
- The paper adopts the synthetic estimate of adaptation needs for Nauru based on the consistency between random forest projections and metadata rankings.

### The Pacific Resilience Facility (PRF): mandate, capitalization, and prospects
- Mandate and governance:
  - Pacific‑led, Pacific‑owned regional financing mechanism to strengthen community‑level resilience.
  - Formally endorsed at the Special Forum Economic Ministers Meeting (FEMM) held in Tonga on March 25–26, 2025; domiciled in Tonga.
  - Intended to complement larger multilateral mechanisms by providing smaller, faster‑disbursing grants targeted at community‑level priorities.
- Capitalization, pledges, and timing:
  - Aim: mobilize an initial US$500 million in grant resources, to be raised in two tranches of US$250 million.
  - To date, fifteen donors have pledged around US$173 million.
  - Japan and Australia have confirmed that their contributions will be transferred to the PRF account at the Pacific Islands Forum Secretariat by August 2027.
  - Current fundraising gap: US$330–343 million relative to the US$500 million target.
  - PRF expected to launch its first call for grant proposals in 2026, aligned with the 55th Pacific Islands Forum Leaders Meeting in Palau.
- Prospects and challenges:
  - Achieving full US$500 million capitalization is critical but uncertain amid competing global financing priorities and fiscal pressures in donor countries.
  - Fundraising challenges and delays in pledged contributions becoming available suggest full capitalization may take time.
  - PRF has generated strong political momentum within the region; success depends on fundraising and timely transfer of pledged contributions.
- Operational rationale:
  - PRF aims to address access, disbursement speed, and alignment with community‑level implementation capacity by pooling grant resources and offering smaller, faster‑disbursing grants.

### Annex highlights
- Annex I: full list of country reports used as sources for the metadata analysis (selected sources listed, e.g., World Bank CCDR 2024 for atoll countries; IMF departmental and technical papers).
- Annex II: sector classification (Table II.1) mapping 56 original sector categories into broader sectors (examples provided).
- Annex II: sources and definitions of variables for Random Forest model (Table II.2) listing data sources such as WEO2019, UNCTAD2019, EM-DAT, Climate-driven INFORM Risk Indicator (2022), IMF-adapted ND-GAIN scores (2021).

*Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026083-source-pdf.pdf*

### Executive Summary ......................................................................................................

### Executive Summary

### Key findings on adaptation needs
- Pacific Island Countries (PICs) face acute and rising climate adaptation needs owing to exposure to sea‑level rise, natural hazards, and structural vulnerabilities associated with small, remote island economies.
- For atoll countries such as Kiribati, Tuvalu, and the Marshall Islands, projected sea‑level rise of up to 0.5 meters by 2070–2110 could submerge large portions of urban areas.
- Physical risks—particularly from coastal inundation and sea‑level rise—account for the bulk of identified needs, with the highest burdens falling on atoll nations.
- When expressed relative to GDP, adaptation needs are exceptionally high in most PICs, reflecting both the magnitude of required investments and the region’s relatively small economic base.
- Existing assessments vary widely in scope, methodology, sectoral coverage, and time horizon, creating challenges for aggregation and comparison.

### Methodological approach
- The paper develops a unified framework combining:
  - A metadata analysis that assembles information from existing country reports and standardizes it into comparable annual measures of adaptation needs.
  - A complementary machine‑learning approach to generate synthetic estimates for data‑deficient countries and to assess internal coherence across metadata‑based estimates.
- The metadata framework converts estimates into annual percent‑of‑GDP and 2024 USD values and aligns time horizons for regional analysis.
- The analysis relies on definitions and scope specified in the underlying source studies; reported estimates are conditional on protection targets, time horizons, climate scenarios, and financing conditions.
- Table 2.1 summary statistics (selected):
  - Fiji: No. of Reports 3; No. of Sectors 14; End of Timeframe 2060
  - Kiribati: No. of Reports 4; No. of Sectors 20; End of Timeframe 2150
  - Marshall Islands: No. of Reports 3; No. of Sectors 7; End of Timeframe 2150
  - Micronesia: No. of Reports 4; No. of Sectors 13; End of Timeframe 2060
  - Nauru: No. of Reports 1; No. of Sectors 1; End of Timeframe 2035 (note: for Nauru, there are a total of 9 sectors, but only 1 extends until 2035, while the others end in 2023)
  - Palau: No. of Reports 3; No. of Sectors 9; End of Timeframe 2060
  - Papua New Guinea: No. of Reports 2; No. of Sectors 5; End of Timeframe 2030
  - Samoa: No. of Reports 3; No. of Sectors 12; End of Timeframe 2060
  - Solomon Islands: No. of Reports 4; No. of Sectors 10; End of Timeframe 2060
  - Tonga: No. of Reports 4; No. of Sectors 15; End of Timeframe 2060
  - Tuvalu: No. of Reports 4; No. of Sectors 13; End of Timeframe 2150
  - Vanuatu: No. of Reports 3; No. of Sectors 28; End of Timeframe 2060

### Adaptation finance flows and delivery
- Climate‑related commitments to the region have increased in recent years, with adaptation as the predominant focus of climate finance providers.
- Metadata analysis indicates current commitment and disbursement levels remain far below projected needs.
- Disbursement ratios vary significantly across financing channels:
  - Generally higher for bilateral partners.
  - Generally lower for multilateral institutions and climate funds.
- Variation in disbursement ratios reflects differences in project design, approval processes, and implementation capacity.

### Determinants of finance mobilization
- Vulnerability to climate change remains a strong determinant of resource allocation.
- Institutional capacity and governance readiness appear to influence financing outcomes.
- Weak capacity continues to constrain implementation in several countries.
- External factors—most notably fiscal pressures in provider countries—may affect the availability of future climate finance.

### Adaptation financing gaps and implications
- Harmonized, region‑wide estimates show large and uneven adaptation needs across PICs.
- Current financing commitments and disbursements are insufficient to meet projected adaptation needs, creating substantial adaptation‑financing gaps.
- Differences in source material, potential overlaps in reported cost components, and extended time horizons in some assessments introduce uncertainty and warrant caution in cross‑country interpretation.

### Policy implications and priorities
- Urgent expansion and acceleration of adaptation finance for Pacific Island Countries are required.
- Key priorities include:
  - Enhancing institutional readiness.
  - Strengthening public financial management for climate‑related investments.
  - Improving the effectiveness and accessibility of financing mechanisms, including emerging regional platforms such as the Pacific Resilience Facility.
- Addressing implementation capacity constraints will be central to improving disbursement performance and closing financing gaps.

### Contributions of the paper
- The paper’s contribution is threefold:
  - It develops a unified framework to organize and standardize climate adaptation needs drawn from diverse country reports, enabling consistent cross‑country comparison despite differences in scope, methodology, and time horizons.
  - It complements the metadata analysis with a machine‑learning approach that helps address gaps in the underlying literature—particularly for data‑deficient countries—and provides an additional perspective to help interpret cross‑country differences.
  - It brings together harmonized adaptation‑needs estimates with detailed evidence on climate finance commitments and disbursements to PICs, offering a new assessment of adaptation financing gaps and factors associated with effective access to climate finance.

*IMF Working Paper — Executive Summary*

### Annex I also provides a full list of country reports used as sources for the metadata analysis.

### wpiea2026083-source-pdf - Annex I also provides a full list of country reports used as sources for the metadata analysis.

### Data sources and literature coverage
- Primary cited analyses used for adaptation financing needs:
  - IMF departmental paper on Fiscal Policy to Address Climate Change in Asia and the Pacific (Dabla-Norris et al. (2021)) — covers most PIC countries except Nauru.
  - World Bank Country Climate and Development Report (CCDR) for the Pacific Atoll Countries (World Bank (2024)) — covers Kiribati, Tuvalu, and the Marshall Islands.
  - World Bank Climate Change and Disaster Management (CCDM) report (2016) — addresses most PICs, except Papua New Guinea and Nauru.
- UNEP’s Adaptation Gap Report (2023) presents aggregated estimates but is not cited directly because it is based on the above country-specific reports.
- Annex I lists country-specific reports used as sources for the metadata analysis.

### Matrix framework and data processing steps
- Purpose: estimate adaptation needs for each country using a standardized matrix to enable regional summation.
- Step 1:
  - Search reports to determine which sectors are assessed for financing needs.
  - Record sector names and exact amounts in original units (percentage of GDP or U.S. dollar values).
  - Enter each sector as a separate row with corresponding values in columns.
- Step 2:
  - Record time span for each estimate and standardize estimates as annual percentage of GDP in one column.
- Step 3:
  - Convert standardized annual cost into 2024 USD millions by:
    - Applying actual U.S. inflation rate to rebase estimates expressed in past years.
    - Applying a 2 percent inflation assumption (U.S. long-run inflation target) to extrapolate estimates expressed in future years.
- Note: The conversion abstracts from uncertainty in future inflation, exchange rates, and financing conditions and is intended to provide order-of-magnitude estimates rather than precise projections.

### Sector classification and aggregation
- Original reports yield 56 sector categories; the framework preserves original terminology (e.g., "education" and "social spending – education" kept separate).
- Categories are pooled into two groups corresponding to the risks they are adapted to.
- Detailed list of sectors and corresponding risks available in Annex II (Table 1).
- For concise assessment, detailed sectors summarized into:
  - Primary sectors: capture largest adaptation needs from direct exposure of physical capital to natural disasters and rising sea levels (e.g., coastal protection infrastructure, land reclamation).
  - Secondary sectors: all other sectors (examples include Agriculture, Tourism Resilience, Biosecurity, Ocean and Marine Protection, Wetland Conservation, Capacity Building, Climate Policy, Education, Health Services, Social Protection, Water and Sanitation).
- Rationale: largest cost estimates in key source reports are concentrated in primary sectors; reclassifying small secondary components would not materially affect aggregate estimates.

### Caveats and methodological limitations
- Potential overlap or double counting of financing needs estimates across sources.
- Possible omission of relevant literature.
- Sources differ in scope and methodology (from incremental resilience upgrades to comprehensive adaptation), so large GDP shares do not imply full climate insulation.
- The cost of reallocating the entire population to a newly established site in another country lies beyond the scope of existing analysis and is not captured.
- Some sectors may be misclassified; however, main cost drivers are in primary sectors.

### Findings: Annual financing needs (metadata analysis)
- Regional total (Pacific Island Countries, PICs):
  - Annual financing need around USD 3.3 billion (at 2024 prices).
  - Physical risks account for 56 percent of needs.
  - Other risks account for 44 percent of needs.
- Country-level annual needs (2024 USD million):
  - Papua New Guinea: USD 868 million.
  - Fiji: USD 857 million.
  - Samoa, Micronesia, Tonga, the Marshall Islands, Kiribati: needs above 100 million and below 200 million.
  - Nauru: only one study available for needs beyond 2023 and covers only other risks.
- Regional needs as percent of GDP:
  - Average annual needs: 22.5 percent of GDP.
  - Physical risks: 13.1 percent of GDP.
  - Other risks: 9.4 percent of GDP.
  - Footnote: This amounts to 24.5 percent of GDP when using the random‑forest–estimated needs for Nauru, as reported in Section 2.2.3.
- Sectoral patterns by country:
  - Physical risks dominate in Tuvalu (97 percent), the Marshall Islands (85 percent), and Kiribati, Solomon Islands, Palau, and Papua New Guinea (all above 60 percent).
  - Other risks dominate in Micronesia, Tonga, and Samoa.

### Time horizons covered by assessments
- Time horizons vary significantly across assessments: range from 10 to 125 years.
- Physical risks:
  - Often projected over much longer periods; for the three atoll countries (Kiribati, the Marshall Islands, Tuvalu) assessments consider a two-meter sea-level rise beyond 2150, yielding a 125-year time span.
  - In non-atoll countries, World Bank CCDM typically uses a 40-year horizon, divided into 2020–2040 and 2040–2060.
- Other risks:
  - Generally assessed over the medium term—up to 2050; many other assessments adopt approximately a 10-year horizon.
- Examples of assessment time spans:
  - IMF Working Paper on Fiscal Policies to Address Climate Change in Asia and the Pacific: 2021–2030.
  - Kiribati Joint Implementation Plan for Climate Change and Disaster Risk Management: 2013–2023.
  - IMF Technical Assistance Report on Climate Change Policy in Tonga: 2018–2028.
  - Fiji’s Climate Vulnerability Assessment: 2017–2027.
- Data gaps:
  - Nauru: most identified financing needs end in 2023, with one analysis on water and sanitation extending to 2035.
  - Papua New Guinea: no financing needs assessment extending beyond 2030.

### Machine learning approach for data‑deficient countries
- Purpose:
  - Address heterogeneity in coverage, timespan, and methodology across literature.
  - Generate synthetic estimates for data‑deficient countries (e.g., Nauru) based on relationships between adaptation needs and country characteristics.
- Model choice and rationale:
  - Random forest model selected for relative robustness to small sample size.
  - Training data: 11 countries (Nauru excluded from training).
  - Due to small sample, no train/test split; visual checks used for relationships between dependent and important independent variables.
  - To address uncertainty, results averaged from 200 random forests (Bayesian spirit).
  - Each forest:
    - Comprises 1000 trees.
    - Each tree trained on 4 independent variables randomly selected from a pool of 24 independent variables.
    - Unlimited tree depth.

### Random forest model inputs and dependent variable
- Dependent variable:
  - Annualized adaptation needs (percentage of GDP, annualized).
  - Table 2.3 summary statistics (as presented):
    - Depedent Var. Adaptaion needs (% of GDP, annualized) 11 24.2 12.5
- Explanatory variables (24 total) span economic indicators, geographic attributes, and climate‑related vulnerability metrics. Examples and summary statistics (as presented) include:
  - GDP per capita (US$): N 11, Mean 12.33, Std. Dev. 3893.8
  - Nominal GDP (US$ mil.): N 11, Mean 3207.47, Std. Dev. 47307.0
  - Public debt (% of GDP): N 11, Mean 36.9, Std. Dev. 26.6
  - Population (thousands): N 11, Mean 35436, Std. Dev. 37
  - Tourism, dummy: N 11, Mean 0.45, Std. Dev. 0.52
  - Distance to economic centers: N 11, Mean 168.7, Std. Dev. 12.3
  - Distance to nearest (non-SIDs) neighbor: N 11, Mean 57.2, Std. Dev. 28.5
  - Distance to trading partners: N 11, Mean 50.5, Std. Dev. 7.8
  - Liner Shipping Connectivity: N 11, Mean 24.5, Std. Dev. 13.8
  - Freight costs of imports (% of good imports): N 11, Mean 14.2, Std. Dev. 5.7
  - Atoll, dummy: N 11, Mean 0.27, Std. Dev. 0.47
  - Disaster Severity (% of population): N 11, Mean 5.2, Std. Dev. 18.2
  - Disasters Frequency, Flood: N 11, Mean 0.14, Std. Dev. 0.20
  - Disasters Frequency, Storm: N 11, Mean 0.27, Std. Dev. 0.24
  - Disasters Frequency, Total: N 11, Mean 0.50, Std. Dev. 0.42
  - Expected Severity (% of population): N 11, Mean 4.55, Std. Dev. 3.14
  - Climate risk: N 11, Mean 3.7, Std. Dev. 0.70
  - Climate risk, Hazard & Exposure: N 11, Mean 2.5, Std. Dev. 0.76
  - Climate risk, Vulnerability: N 11, Mean 4.1, Std. Dev. 0.60
  - Lack of Coping capacity: N 11, Mean 5.0, Std. Dev. 1.30
  - Readiness: N 11, Mean 0.43, Std. Dev. 0.09
  - Economic Readiness: N 10, Mean 0.44, Std. Dev. 0.14
  - Vunerability: N 9, Mean 0.55, Std. Dev. 0.06
  - Exposure: N 11, Mean 0.54, Std. Dev. 0.06
- Note: Table AII.2 in Annex II provides sources and definitions of variables.

### Machine learning training results and interpretive findings
- Overall:
  - Adaptation needs are significantly correlated with most explanatory variables.
  - Figure 2.6 identifies relative importance of explanatory variables; eight variables stand out, while others remain relatively important.
- Key observed relationships (visual assessment):
  - Nominal GDP: strong negative relationship with adaptation needs (panel 1) — smaller economies exhibit higher required adaptation investments per GDP.
  - Public debt: negative correlation with adaptation needs (panel 2) — low public debt may reflect limited borrowing capacity and higher unmet needs.
  - Geographic remoteness and higher shipping costs: positive association with higher adaptation investment costs (panels 3–4).
  - Climate-related vulnerabilities (e.g., exposure to natural disasters): positive relationship with adaptation needs (panel 5).
  - Atoll countries (highlighted in orange in Figure 2.7) tend to face substantial adaptation needs, reflecting fragility driven by small size, remoteness, high shipping costs, and greater exposure to disasters.
- Interpretation:
  - Expected associations from the random forest visual checks support the plausibility of metadata-derived estimates despite heterogeneity in source coverage, time horizons, and methodologies.
- Uncertainty and limitations:
  - The prominence of many variables may partly reflect estimation uncertainty given small sample size.
  - For Nauru, large uncertainty in metadata estimates led to omission of the metadata estimate; the machine learning approach generates a synthetic estimate for data-deficient cases.

*Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026083-source-pdf.pdf*

### 1. Economic Size2. Financing Capacity

### 1. Economic Size2. Financing Capacity

### Prediction of Missing Annual Adaptation Needs
- Methodology:
  - Estimated random forest model used to predict annual adaptation needs for 2024 for countries not in the training dataset (example: Nauru).
  - Alternative approach: use synthetic estimates for all countries as baseline (random-forest shrinkage toward central tendencies).
- Key quantitative findings (Figure 2.8, Imputed Adaptation Needs, In percent of GDP, annualized, estimated annual requirement for 2024):
  - 35.7
  - 34.3
  - 32.7
  - 30.1
  - 28.3
  - 27.9
  - 23.6
  - 23.0
  - 21.3
  - 17.1
  - 14.7
  - 11.5
- Country-specific insight:
  - Nauru (NRU) is predicted to require the fifth-largest adaptation investment as a percentage of GDP among the 12 PICs (Figure 2.8, orange).
  - The random forest projection (RF-Projection) preserves the rankings from the metadata analysis (Meta Analysis), with mid-ranked countries showing minimal discrepancies between methods.
- Implication:
  - The paper adopts the synthetic estimate of adaptation needs for Nauru based on the consistency between random forest projections and metadata rankings.

### Assessing Actual Adaptation Finance Flows to Pacific Island Countries
- Objective and scope:
  - Assess adaptation finance flows committed to PICs and effectively disbursed over the past decade by bridging existing data collection on climate finance flows and highlighting key drivers for successful mobilization.
  - Time coverage for commitment data: 2010-2023, compiled with an approximately 1.5-year lag, based on self-reporting from multilateral, bilateral and private sources.
- Data sources and limitations:
  - Commitment data formally reported to the UNFCCC and collected by the OECD-DAC, integrated into ODA tracking.
  - Bilateral data rely on “Rio markers” (identify projects with principal or significant climate objectives).
  - Multilateral agencies primarily use the Joint MDB Methodology on Tracking Climate Finance (identifies shares supporting adaptation or mitigation).
  - Caveat: data are self-reported and rely on tracking methodologies with limitations; commitment data likely to be inflated.
- Global and regional headline figures:
  - Global climate flows reached USD 136 billion in 2022.
  - Out of that total, USD 65 billion were commitments with adaptation as a principal or significant objective.
  - Paris agreement commitment: US$ 100 billion annual goal (developed nations to mobilize by year 2020).
- Adaptation finance in PICs:
  - In 2023, adaptation was the objective in 90 percent of all climate finance commitments to PICs.
  - PICs received about 2 percent of global adaptation finance commitments over the 2019-2023 period.
  - PNG accounted for a third of all adaptation finance commitments to PICs and is noted as representing 73 percent of PICs’ GDP and 79 percent of PICs’ population.
- Temporal pattern:
  - Adaptation finance commitments accelerated over the past decade: more commitments over the 2019-2023 period than over the eight years prior.
  - 2022 marked a milestone as global climate finance targets were achieved and surpassed.
- Composition by instrument (2011-2023):
  - Grants represented 84 percent of total adaptation finance commitments.
  - Half of PICs relied solely or almost solely on grants.
  - The four countries that relied the most on loans are the four larger economies in the region.
- Composition by source (2011-2023):
  - 60 percent of adaptation finance commitments to the PIC region came from bilateral donors.
  - 40 percent came from multilateral donors.
  - Multilateral composition (share of commitments):
    - World Bank: 46%
    - ADB: 19%
    - Green Climate Fund: 18%
    - Other Climate Funds: 15%
    - Other MDBs: 2%
  - Finance from private foundations or nonprofit organizations represented a very limited share.
- Country-level commitments (Figure 3.3, Adaptation-Related Development Finance to Individual Pacific Island Countries, Commitments, 2023 USD bn; estimates exclude regional projects):
  - Numerical values shown in the figure (2023 USD billion): 2.15, 0.99, 0.84, 0.75, 0.57, 0.41, 0.36, 0.30, 0.25, 0.23, 0.14, 0.07
- Disbursements versus commitments:
  - Disbursements correspond to actual payments and are preferred from the recipient standpoint.
  - Disbursements are often tranches of multiyear projects signed in earlier years; commitments are recorded at signature and reflect full project cost.
  - Rapid increases in commitments typically translate into higher disbursements only after a few years.
  - Estimating actual disbursements is challenging due to lack of a comprehensive official reporting framework; initiatives by non-governmental bodies exist but data limitations remain.
- Data reporting notes:
  - OECD tasked in 2015 to track progress against the USD 100 billion goal, though OECD does not have a formal climate finance monitoring role under the UNFCCC.
  - Tracking methodologies (Rio markers, Joint MDB Methodology) are useful but have limitations (e.g., projects with only limited focus on climate).

*IMF WORKING PAPERS — Climate Finance and Adaptation Needs In Pacific Island Countries*

### Box 1. The Pacific Resilience Facility: Mandate, Capitalization, and Prospects

### Box 1. The Pacific Resilience Facility: Mandate, Capitalization, and Prospects

### Mandate and governance
- The Pacific Resilience Facility (PRF) is a Pacific‑led, Pacific‑owned regional financing mechanism designed to strengthen community‑level resilience to climate change and natural disasters.
- Formally endorsed by Pacific Island Forum members at the Special Forum Economic Ministers Meeting (FEMM) held in Tonga on March 25–26, 2025, establishing the Facility through a treaty‑based framework.
- The PRF will be domiciled in Tonga.
- The PRF seeks to pool resources under a single Pacific‑governed platform to provide smaller, faster‑disbursing grants targeted at locally identified resilience priorities, particularly at the community level.
- The PRF is intended to complement—rather than duplicate—larger multilateral mechanisms whose procedures and project sizes often do not match the needs and implementation capacity of small island communities.

### Capitalization, pledges, and timing
- The Facility aims to mobilize an initial US$500 million in grant resources, to be raised in two tranches of US$250 million.
- To date, fifteen donors have pledged around US$173 million.
- Japan and Australia have confirmed that their contributions will be transferred to the PRF account at the Pacific Islands Forum Secretariat by August 2027.
- The current fundraising gap is US$330–343 million relative to the US$500 million target.
- The PRF is expected to launch its first call for grant proposals in 2026, aligned with the 55th Pacific Islands Forum Leaders Meeting in Palau.

### Prospects, challenges, and political momentum
- Achieving full US$500 million capitalization will be critical to the Facility’s overall firepower but is uncertain amid competing global financing priorities and fiscal pressures in donor countries.
- Fundraising challenges identified by Pacific authorities, and delays in pledged contributions becoming available for disbursement, suggest that achieving full capitalization may take time.
- The Facility has generated strong political momentum within the region, with leaders emphasizing local ownership and predictable access to adaptation resources as central to the Pacific’s long‑term resilience strategy.

### Operational rationale and expected role
- Pacific authorities view global climate finance as increasingly scarce, fragmented, and difficult to access, with slow disbursement and complex application processes.
- By pooling grant resources and offering smaller, faster‑disbursing grants, the PRF aims to address access, disbursement speed, and alignment with community‑level implementation capacity.
- The PRF’s design focuses on locally identified resilience priorities, seeking to improve predictability and accessibility of adaptation resources for small island communities.

### Key implications highlighted in the source
- Continued engagement with development partners is necessary to close the US$330–343 million capitalization gap.
- The PRF’s success depends on both fundraising and timely transfer of pledged contributions (examples: Japan and Australia commitments by August 2027).
- Strengthening institutional readiness, project implementation capacity, and effectiveness of regional mechanisms (including the PRF) is central to boosting resilience in Pacific Island Countries.

*Box 1. The Pacific Resilience Facility: Mandate, Capitalization, and Prospects.*

### Annex I. List of Country Report Sources Used In

### Annex I. List of Country Report Sources Used In Metadata Analysis

### Country report sources
- Kiribati — World Bank's Pacific Atoll Countries CCDR (2024) 4
  - World Bank’s Climate Change and Disaster Management (2016)
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Kiribati's Joint Implementation Plan on Climate Change and Disaster Risk Management (2014)
- Micronesia — World Bank’s Climate Change and Disaster Management (2016) 4
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Micronesia's Climate Change Policy Assessment from the IMF (2017)
  - Readiness Proposal GCF 2023
- Solomon Islands — World Bank’s Climate Change and Disaster Management (2016) 4
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Solomon Islands' NDC (2021)
  - IMF Selected Issues Paper 2022 on Spending Needs for Achieving SDGS with Climate Resilience
- Tonga — World Bank’s Climate Change and Disaster Management (2016) 4
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Tonga's Joint Implementation Plan on Climate Change and Disaster Risk Management 2018-2028 (2018)
  - IMF Technical Assistance Report on Climate Change Policy Assessment for Tonga (2020)
- Tuvalu — World Bank's Pacific Atoll Countries CCDR (2024) 4
  - World Bank’s Climate Change and Disaster Management (2016)
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Tuvalu’s National Strategic Action Plan: 2012-2016
- Fiji — World Bank’s Climate Change and Disaster Management (2016) 3
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Fiji's Climate Vulnerability Assessment (2017)
- Marshall Islands — World Bank's Pacific Atoll Countries CCDR (2024) 3
  - World Bank’s Climate Change and Disaster Management (2016)
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
- Palau — World Bank’s Climate Change and Disaster Management (2016) 3
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Palau's Selected Issues on Climate Change by the IMF (2023)
- Samoa — World Bank’s Climate Change and Disaster Management (2016) 3
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Samoa's Technical Assistance Report on Climate Macroeconomic Assessment Program from the IMF (2022)
- Vanuatu — World Bank’s Climate Change and Disaster Management (2016) 3
  - IMF Working Paper on Fiscal Policy to Address Climate Change (2021)
  - Vanuatu's NDC (2021)
- Papua New Guinea — IMF Working Paper on Fiscal Policy to Address Climate Change (2021) 2
  - Papua New Guinea's Second NDC (2020)
- Nauru — Nauru's NDC (2022) 1

### Annex II. Sector classification (Table II.1)
- Sector — Corresponding Broader Sector
  - Coastal Erosion Control — Coastal Protection
  - Coastal Protection Infrastructure — Hazard Management
  - Inundation and Flood Mitigation — Physical Risks
  - Meteorology and Climate Information Services — Protect and Reclaim Land, and Internal Migration
  - Protect Land, Raising Building, and Move Inland — Storm Impact Mitigation (Capital)
  - Agriculture — Building and Housing
  - Energy Infrastructure Upgrades — Financial Services and Climate Finance
  - Fisheries Management — Fisheries Revenue Loss Mitigation
  - ICT (Information and Communication Technology) — Infrastructure Development
  - Livestock Management — Maintenance
  - Pooled Funding Mechanisms — Private Sector Greening
  - Renewable Energy Promotion — Risk Management
  - Tourism Resilience — Transport Infrastructure
  - Biosecurity — Environmental Protection
  - Forestry — Land Management
  - Ocean and Marine Protection — Wetland Conservation
  - Capacity Building — Other Sectors
  - Climate Policy — Contingency Planning
  - Decentralization — Disability Inclusion
  - Early Warning Systems — Education
  - Gender Equality — Governance
  - Health Services — Human Rights Advocacy
  - Indigenous Rights — Knowledge and Information Sharing
  - Planning and Policy Development — Population Relocation
  - Social Protection — Social Spending - Education
  - Social Spending - Health — Social Spending - Health and Education
  - Sovereignty Maintenance — Storm Impact Mitigation (Population)
  - Youth Engagement — Waste Management
  - Water and Food Security — Water and Sanitation

- Source: authors.

### Sources and definitions of variables for Random Forest model (Table II.2)
- Economic
  - GDP per capita — WEO2019
  - Nominal GDP — WEO2019
  - Public debt — WEO Public debt to GDP ratio (2023)
  - Population — WEO2019
  - Tourism — APD REO 1=tourism country
- Geographic
  - Distance to economic centers — UNCTAD2019
  - Distance to nearest (non-SIDs) neighbor — UNCTAD2019
  - Distance to trading partners — UNCTAD2019
  - Liner Shipping Connectivity — UNCTAD2024Q4
  - Freight costs of imports — Staff Calculation Percent of good imports (2018)
  - Atoll — World Bank CCDR 1=atoll country
- Vulnerability
  - Disaster Severity — EM-DAT Average Affected Population Share (2000-2023)
  - Disasters Frequency, Flood — EM-DAT 2000-2023
  - Disasters Frequency, Storm — EM-DAT 2000-2023
  - Disasters Frequency, Total — EM-DAT 2000-2023
  - Expected Severity — EM-DAT Severity X Disaster Frequency (2000-2023)
- Climate risk
  - Climate-driven INFORM Risk Indicator — Climate-driven INFORM Risk Indicator (2022)
  - Climate risk, Hazard & Exposure — Climate-driven INFORM Risk Indicator Sub-index for Hazard & Exposure (2022)
  - Climate risk, Vulnerability — Climate-driven INFORM Risk Indicator Sub-index for vulnerability (2022)
  - Lack of Coping capacity — Climate-driven INFORM Risk Indicator Sub-index for coping capacity (2022)
- Readiness and vulnerability indices
  - Readiness — IMF-adapted ND-GAIN index Readiness score (2021)
  - Economic Readiness — IMF-adapted ND-GAIN index Readiness score sub-index, Economic (2021)
  - Vunerability — IMF-adapted ND-GAIN index Vunerability score (2021)
  - Exposure — IMF-adapted ND-GAIN index Vulnerability score sub-index, Exposure (2021)

### Select references cited within the annex (excerpt)
- Bellon, M., Z. Aligishiev, and E. Massetti. 2022. “Macro‑Fiscal Implications of Adaptation to Climate Change.” IMF Staff Climate Note No. 2022/002. Washington, DC: International Monetary Fund. https://doi.org/10.5089/9798400201608.066
- Bellon, M. and E. Massetti. 2022a. “Economic Principles for Integrating Adaptation to Climate Change into Fiscal Policy.” IMF Staff Climate Note No. 2022/001. Washington, DC: International Monetary Fund. https://www.imf.org/en/Publications/staff-climate
- Bellon, M. and E. Massetti. 2022b. “Planning and Mainstreaming Adaptation to Climate Change in Fiscal Policy.” IMF Staff Climate Note No. 2022/003. Washington, DC: International Monetary Fund. https://doi.org/10.5089/9798400201950.066
- Betzold, C. and F. Weiler. 2017. “Allocation of aid for adaptation to climate change: Do vulnerable countries receive more support?”. International Environmental Agreements Vol. 17, pp. 17–36. https://doi.org/10.1007/s10784-016-9343-8
- Dabla-Norris E., J. Daniel, M. Nozaki, C. Alonso, V. Balasundharam, M. Bellon, C. Chen, D. Corvino and J. Kilpatrick. 2021. "Fiscal Policies to Address Climate Change in Asia and the Pacific." IMF Departmental Paper 21/07. Washington, DC: IMF.
- Fouad M., N. Novta, G. Preston, T. Schneider and S. Weerathunga. 2021. “Unlocking Access to Climate Finance for Pacific Island Countries” IMF Departmental Paper 21/20. Washington, DC: IMF.
- Government of Kiribati. 2014. "The Kiribati Joint Implementation Plan on Climate Change and Disaster Risk Management (KJIP)." Presented in Intended Nationally Determined Contribution. Tarawa, Kiribati.
- Government of Nauru. 2022. "Nationally Determined Contribution (NDC)." Yaren, Nauru.
- Government of Papua New Guinea. 2020. "Second Nationally Determined Contribution (NDC)." Port Moresby, Papua New Guinea.
- Government of the Federated States of Micronesia. 2019. "Climate Change Policy Assessment." Palikir, Federated States of Micronesia.
- Government of Solomon Islands. 2021. "Nationally Determined Contribution (NDC) Report.". Honiara, Solomon Islands.
- Green Climate Fund. 2023. "Micronesia: Readiness Proposal." Songdo, Republic of Korea.
- Hallegatte, S., J. Rentschler and J. Rozenberg. 2020. Adaptation Principles: A Guide for Designing Strategies for Climate Change Adaptation and Resilience. Washington, DC: World Bank.
- Intergovernmental Panel on Climate Change. 2022. IPCC 6th Assessment Report. Geneva, Switzerland.
- World Bank Group. 2024. "Country Climate and Development Report (CCDR) for the Pacific Atoll Countries." Washington, DC: World Bank Group.
- World Bank Group. 2016. "Climate Change and Disaster Management." 2016. Washington, DC: World Bank Group.

*Climate Finance and Adaptation Needs In Pacific Island Countries, Working Paper No. WP/2026/083. INTERNATIONAL MONETARY FUND.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026083-source-pdf.pdf_