Keywords

1 Introduction

The global community continues to be at risk of landslide hazards. Although the Emergency Events Database (EM- DAT) has shown a decline in annual landslides for the last two years, compared to the 20-year annual average (CRED 2022, 2023), other databases that record landslide events have indicated that the frequency and impact of landslides remains higher than what the EM-DAT database reports. Landslide events and subsequent losses incurred from these events are often underreported. This is likely because landslides are typically a secondary hazard that is triggered as a consequence of another hazard (Froude and Petley 2018). Between 1980 and 2018, there was an annual average of 400,000 rainfall-triggered landslides and 130,000 earthquake-triggered landslides (The World Bank 2021). According to the same report, there was an observed 1-2 percent increase in the average rainfall-triggered landslides each year, which could be indicative of a broader effect of climate change on a global scale. Additionally, according to several landslide databases, there is also an increasing percentage of landslides occurring from anthropogenic causes such as illegal mining and construction activities (Gómez et al. 2023). The EM-DAT database is also limited to landslide events that are considered disasters by their definition and it encompasses reports from only 82 countries. In contrast, other databases have a broader coverage and have recorded significantly more landslide events. While EM-DAT has recorded only 640 landslide disasters from 1903 to 2020 and 50,144 human losses, the global fatal landslide database has collected information on 5490 landslides from 2004 to 2017 corresponding to 64,218 human losses. The Global Landslide Catalog database has recorded 6784 landslides between 2007 and 2015 with a total of 25,386 human losses. Finally, the Desinventar database contains 25,195 unique records of landslides from 1906 to 2020 corresponding to 46,724 human losses. With these gaps in reporting and underestimations of occurrences and impacts, the global community must continue their efforts in addressing landslide risks through robust risk governance and management, and collaboration among various stakeholders.

To prevent disasters and reduce the risk of hazards, including landslides, the international community has committed to ambitious goals set out in the Sendai Framework for Disaster Risk Reduction 2015–2030. The Sendai Framework is a global agreement adopted by the United Nations (UN) General Assembly (GA) in 2015 at the Third UN World Conference on Disaster Risk Reduction (DRR) held in Sendai, Japan. The framework aims to reduce risks by advocating a comprehensive and all-of-society approach. It sets out seven targets and four priorities for action to guide countries and stakeholders in achieving substantial reductions in disasters and disaster losses. By emphasizing risk reduction rather than just response and recovery, the Sendai Framework seeks to create more resilient communities whilst promoting sustainable development in the face of increasing risks posed by natural and man-made hazards, such as that of landslides.

This year marks the midterm of the agreed duration of the framework and a review has been conducted by the United Nations Office for Disaster Risk Reduction (UNDRR) to share the progress made through global collective efforts. The Report of the Midterm Review (MTR) of the Implementation of the Sendai Framework for Disaster Risk Reduction 2015–2030 has concluded that although some progress has been made towards the expected outcomes and goals of the framework, it recognizes the increasing difficulties and challenges we face such as the COVID-19 pandemic, climate change and population growth which calls for course corrections to continue on a positive path towards disaster risk reduction by addressing these interconnected and systemic risks (UNDRR 2023). As UNDRR continues to advocate for an all-of-society approach, the MTR encourages increased participation and better coordination among all stakeholders for more impactful and effective interventions, including non-state actors. UNDRR recognizes that non-state actors, including non- governmental organizations, private sector, academic and research institutions, and media, have an important role in influencing the outcomes by 2030. The MTR has identified challenges and opportunities that can be addressed and supported by non-state actors including but not limited to: strengthening capacities for developing, collecting, analyzing and interpreting disaster risk data whilst improving the standard of official risk data and broadening the application of risk assessments; data sharing of risk information to and from non-state actors among government agencies and each other; and aligning digital innovations with local needs and robust engagement mechanisms for planning and delivery by national and local actors.

Since 2018, UNDRR has been monitoring and taking stock of non-state actors’ contributions to the Sendai Framework through the Sendai Framework Voluntary Commitments (SFVC) online platform. The SFVC online platform was established in response to GA resolution A/69/283 (2015) calling for all stakeholders to share voluntary commitments (VCs) that are specific and time bound to address disaster risks at local, national, regional, and global levels. The SFVC online platform encourages and promotes voluntary actions by various stakeholders towards achieving the objectives of the Sendai Framework for Disaster Risk Reduction. It serves as a repository for commitments made by non-governmental organizations, businesses, civil society groups, networks, media as well as local governments and international organizations, showcasing their efforts to reduce disaster risks and enhance resilience. The platform allows stakeholders to share their commitments to the public, while also allowing them to monitor and report progress of their commitments. By facilitating transparency and accountability, the platform fosters collaboration, knowledge-sharing, and collective action in addressing disaster risks at local, national, regional, and global levels. The VCs submitted through this platform contribute to the implementation of the Sendai Framework, promoting a more resilient and sustainable future.

This article builds on a previous study conducted in 2020 on the SFVC online platform which showed preliminary results of 37 VCs covering landslides. UNDRR has continued to support stakeholders to submit their commitments so that their contribution to the Sendai Framework targets may be recognized. In so doing, the SFVC online platform has published a total of 116 VCs as of 31 March 2023, which has more than tripled since the previous study. This study aims to better understand VCs covering landslides through a comparative analysis of global trends and landslide VCs, landslide VCs and all published VCs, while determining gaps and opportunities that the SFVC online platform can address to support implementation of the Sendai Framework and reduce the risk of landslides.

2 Data and Methods

The source of the data used for this article is the Sendai Framework Voluntary Commitments (SFVC) online platform operated by UNDRR, specifically all published voluntary commitments (VCs) as of 31 March 2023. This dataset includes 116 published voluntary commitments involving a total of 670 organizations from around the world. Published VCs are stakeholders’ submissions from around the world that are reviewed, validated, and collated by UNDRR to take stock of the non-state actors’ contributions to the Sendai Framework. Each commitment reports various data points to verify the project’s contribution to the Sendai Framework.

This article uses data from the SFVC online platform on variables including hazards, themes, Sendai priorities, targets, and indicators among others, which are further elaborated. Hazards are defined as “a process, phenomenon or human activity that may cause loss of life, injury or other health impacts, property damage, social and economic disruption or environmental degradation” (UNDRR, n.d.). Stakeholders may report that their commitments cover more than one hazard at a time. The SFVC online platform encompasses 17 hazards including landslides. Themes and issues (themes) are DRR-related topics that have been identified and standardized by the UNDRR aiming to be both wide and cross-cutting. The SFVC online platform covers 33 themes and issues and, similar to hazards, stakeholders may report the coverage of more than one theme at a time. The Sendai Framework outlines four priorities for action: Understanding disaster risk; Strengthening disaster risk governance to manage disaster risk; Investing in disaster risk reduction for resilience; and Enhancing disaster preparedness for effective response and to “Build Back Better” in recovery, rehabilitation, and reconstruction. In addition, seven targets and 38 indicators were identified to provide clear actionable goals for stakeholders and to measure the progress towards these goals. Priorities, targets, and indicators are also collected from each commitment,

allowing stakeholders to select multiple to better present their contribution to the implementation of the Sendai Framework. To promote coherence and alignment with the overarching sustainable development agenda, the platform also requests commitments to report the relevant Sustainable Development Goals (SDGs) that their project contributes to. Each commitment must be linked with an official entity that acts as the implementing organization (implementers). A commitment may have more than one implementer and may or may not have partner organizations (other partners) that also support the project, although partners are not necessarily directly involved in implementation. An organization can be involved as implementer or partner in more than one VC. Additionally, donor information is also collected (donors). The start of the commitment and the end of the commitment is also required for each commitment and are used to determine the duration of a commitment. Specific deliverables are also mandatory as a means of verification to showcase the outcomes of the commitment and stakeholders are requested to indicate completed deliverables within their commitments. Multiple deliverables are encouraged. Finally, progress reports are requested every 6 months from the last update of a commitment to ensure accountability.

Table 1 presents various variables and statistical measures for two datasets: VCs (all) and VCs (landslides). Insights were derived from this data and are found in the results section. Other variables were also used in the analysis such as geographical scope and country coverage.

Table 1 Summary statistics (selected variables)

The methodology for conducting analysis on these variables included a data cleaning process and a statistical descriptive analysis. As seen in Table 1, out of 116 VCs, 74 address landslides as a hazard being addressed by their commitment, which is 64 percent of the whole dataset. A review of relevant literature and databases was carried out on landslide disasters and events and this information was likened to the data from the SFVC online platform. A comparative analysis was conducted between the dataset for all published VCs and the dataset for VCs that cover landslides. Additionally, a qualitative review and quantitative text analysis of the description and deliverables of landslide VCs was undertaken.

3 Results

This section will further elaborate on the variables set out in the methodology to better understand VCs working on landslides. This includes understanding voluntary commitments overall through a comparative analysis of two datasets and looking deeper into landslide VCs.

4 Comparing Landslide-Related Initiatives with All Voluntary Commitments

The SFVC online platform presents information on VCs contributed by non-state actors. Out of 116 VCs, there are 74 VCs that address landslides as a hazard. Based on the basic characteristics of all VCs and VCs covering landslides, some differences and similarities can be derived (Fig. 1a, b).

Fig. 1
Two bar graphs labeled A and B compare the basic characteristics with significant and insignificant differences for all V Cs and landslides V Cs. Themes have the highest value in graph A, and deliverables exhibit the highest value in graph B.

(a) Basic characteristics with significant differences. (b) Basic characteristics with insignificant differences. Source: UNDRR Voluntary Commitments

Among the VCs covering landslides, the average number of hazards addressed in each commitment is higher (9.46) compared to the average for all VCs (7.33). This suggests that landslide VCs tend to encompass a broader range of hazards. The themes covered by landslide VCs have a slightly higher average (10.24) compared to all VCs (9.41), which may indicate that efforts targeting landslides often address more thematic areas. Landslide VCs have slightly higher averages for both targets (3.64) and indicators (6.01) compared to all VCs (3.30 and 5.59, respectively). This could imply a more detailed and comprehensive approach to setting targets and monitoring progress in landslide-related commitments. Landslide VCs also have a slightly higher average of partners involved (6.93) for each commitment compared to all VCs (5.53). This may suggest a relatively broader range of stakeholders involved in landslide-related initiatives. Finally, landslide VCs have a slightly longer average duration (9.93) compared to all VCs (8.79) and this may indicate that implementers are carrying out activities with the long-term outcomes in mind.

The average for other basic characteristics of landslide VCs is similar to that of all VCs including priorities for action, Sustainable Development Goals (SDGs), and implementers, among others (Fig. 1b). However, we did observe that donor involvement appears to be minimal in both types of commitments with an average of around 1 for both landslide VCs and all VCs. Additionally, the average number of submitted progress reports for both types of commitments is alarmingly low (less than 1), indicating that there is a need to encourage stakeholders to report on progress regarding their commitments.

Besides the basic characteristics above, we also investigated the secured budget and the status of completion for both datasets. While 59 percent of all VCs had fully secured their budget, there was a lower percentage of landslide VCs that had fully secured their budget at 54 percent. The same can be said about the status of completion where 32 percent of all VCs have already been completed, while only 30 percent of landslide VCs were completed. Both observations may be related to the longer duration of landslide VCs compared to all VCs.

Taking a closer look at the hazard coverage for both datasets, we can see that landslides are the third most covered hazard out of 116 VCs, while flood (88 percent) and earthquake (78 percent) are the first and second most covered type of hazard, respectively (Fig. 2a). Comparing the hazard coverage for landslide VCs, we see some changes in the ranking of the hazards, although the difference is not substantial (Fig. 2b). Landslides are often triggered by floods, earthquakes, cyclones, and volcanic eruption, among other hazards, which may likely be the reason for landslides being one of the most covered hazards.

Fig. 2
Two bar graphs exhibit the percentage of the hazards covered by all V Cs and landslides V Cs. Flood has the highest value of 88% for all V Cs, while landslide and flood exhibit the highest value of 100% for landslides V Cs.

(a) Hazards covered by VCs (all) (b) Hazards covered by VCs (landslides). Source: UNDRR Voluntary Commitments

Notes: Epidemic and… stands for Epidemic and pandemic. NBC stands for Nuclear, Biological and Chemical hazards.

The SFVC online platform also collects information on the geographical scope of the VCs, which can be global, regional, or national and local (Fig. 3). Comparing the two datasets, VCs covering all hazards have a more skewed distribution in terms of scope against VCs covering landslides. For landslide-related VCs, more than 50 percent of commitments have a global and regional scope, which suggests that VCs working on landslides may be taking a more comprehensive approach sharing knowledge and information across regions and among countries. The platform also requests information on regional and country coverage which informs Figure 4. There are no observed significant differences in the regional distribution, but it is worth noting that Asia is covered by more than half of the VCs for both datasets, excluding VCs with a global scope.

Fig. 3
A bar graph depicts the scope types for all V Cs and landslides V Cs. For all V Cs, global scope accounts for 28%, regional scope for 18%, and national or local scope for 54%. For landslides V Cs, global scope is 32%, regional scope is 24%, and national or local scope is 43%.

Scope types for VCs (all) and VCs (landslides)

Fig. 4
A bar graph presents the regions for all V Cs and landslides V Cs. For all V Cs, Africa constitutes 16%, the Americas, 13%, Asia, 54%, Europe, 7%, and Oceania, 10%. For landslides V Cs, Africa comprises 15%, the Americas, 12%, Asia, 55%, Europe, 7%, and Oceania, 12%.

Regions for VCs (all) and VCs (landslides). Source: UNDRR Voluntary Commitments

Examining the selected priorities for action for the two datasets, we can see that all priorities have increased coverage when only considering VCs (landslides) except for priority 4 on enhancing disaster preparedness for effective response and to “Build Back Better” in recovery, rehabilitation, and reconstruction (Fig. 5). As for the targets, all targets have higher coverages for VCs covering landslides compared to all published VCs (Fig. 6). We observed particularly higher coverage for target C (7 percent difference), target D (6 percent difference) and target F (8 percent difference) among VCs covering landslides. These targets are relevant to reducing the risk of landslides, as landslides often cause significant damage to critical infrastructure, such as roads, bridges, and utilities, leading to disruptions in essential services. This may also result in cascading effects that disrupt livelihoods and hinder disaster response resulting in more significant economic losses. Finally, developing countries would benefit from international cooperation to enhance landslide risk reduction efforts. Because developing countries must deal with bigger trade-offs between economic development and safety, their risk acceptance for landslides are often more lenient, thereby creating higher societal risk as they focus on activities that would boost economic growth such as job creation, infrastructure development and improved access (Sim et al. 2022). Developed countries can provide support through technical expertise and knowledge sharing, capacity building and training, and financial aid and resources to help developing countries mitigate landslide hazards.

Fig. 5
A bar graph exhibits the priorities for action for all V Cs and landslides V Cs. For all V Cs, priority 1 constitutes 85%, priority 2 is 59%, priority 3 is 54%, and priority 4 is 75%. For landslides V Cs, priority 1 comprises 88%, priority 2 is 64%, priority 3 is 59%, and priority 4 is 72%.

Priorities for action for VCs (all) and VCs

Fig. 6
A bar graph exhibits the Sendai Framework targets for all V Cs and landslides V Cs. All V Cs, A, 47%, B, 79%, C, 50%, D, 33%, E, 51%, F, 30%, and G, 41%. Landslides V Cs, A, 51%, B, 81%, C, 57%, D, 39%, E, 55%, F, 38%, and G, 42%.

Sendai Framework targets for VCs (all) and VCs (landslides) (landslides). Source: UNDRR Voluntary Commitments

The SFVC online platform requests focal points of VCs to select specific themes and issues related to DRR that are applicable to their respective commitments. Delving into these themes for all VCs and landslide VCs, we observed that the ranking of themes and issues based on percentage coverage of VCs remained the same between the two datasets with the top three being disaster risk management, capacity development, and risk identification assessment (Fig. 7). However, significant differences were found in the percentages of the following themes: disaster risk management, risk identification and assessment, vulnerable populations, governance, urban risk and planning, early warning, private sector, water, and economics of DRR. The highest discrepancies were found in risk identification and assessment, vulnerable populations, and private sector, where non-state actors seem to put more focus on such themes and issues in landslide VCs compared to all VCs.

Fig. 7
A double-bar graph illustrates the percentages for themes and issues for all V Cs and landslides V Cs. Disaster risk management exhibits the highest value of 83% for all V Cs and 86% for landslides V Cs. Values are estimated.

Themes and issues for VCs (all) and VCs (landslides). Source: UNDRR Voluntary Commitments

Risk identification and assessment is particularly important for landslide mitigation to help determine areas prone to landslides, enabling targeted mitigation measures and early warning systems. There are many approaches or methods for landslide hazard assessments such as multivariate techniques, physically based models, and multi- criteria decision-making approach (Pardeshi et al. 2013). Vulnerable populations, such as those living in informal settlements or areas with inadequate infrastructure, are at higher risk of landslides. Proper mitigation and response measures like early warning systems, evacuation plans, and slope stabilization can significantly reduce the risk of casualties and injuries and may also reduce the impact on their livelihoods. Finally, the private sector bears a significant stake in the mitigation of landslide risks, as such hazards can potentially lead to supply chain disruption, property damage, and business interruption due to suspension of electricity and water supply, resulting in significant economic losses. The participation of the private companies in landslide risk reduction would be beneficial for both the private sector and public sector and there is great potential for public-private partnerships to increase investments in landslide risk assessments and preventive measures to support vulnerable populations.

As mentioned previously, the average number of selected indicators by landslide VCs is higher than the average number of selected indicators for all VCs. Thus, the percentage of VCs covered by each indicator is mostly higher if note that same. However, there are 9 out of 38 exceptions, namely B2, B3, B4, B5, E1, G3, G5 and G6. However, it should be emphasized that target B and target G both have a compound indicator which is often selected by stakeholders instead of non-compound indicators (Fig. 8).

Fig. 8
A double-bar graph illustrates the percentages of indicators selected for all V Cs and landslides V Cs. B 1 exhibits the highest value of 65% for all V Cs and 69% for landslides V Cs. Values are estimated.

Indicators selected for VCs (all) and VCs (landslide). Source: UNDRR Voluntary Commitments

While other variables from the two datasets were also compared, no other significant variances were observed.

5 A Closer Look at Landslide-Related VCs

As mentioned in previous sections, VCs covering landslides make up 64 percent of all commitments in the SFVC online platform. They have a mix of geographical scope including 24 VCs with global scope, 18 VCs with a regional scope, and 32 VCs with a national or local scope. It is important to note that regional and national or local VCs may cover multiple regions and countries and UNDRR has distinguished five regions for their work. Landslide VCs have a wide coverage of regions and countries involving all five regions and 34 countries in total with a much larger concentration within the Asia region (Fig. 9).

Fig. 9
A world map presents the regions with countries and territories having V Cs that cover landslide hazards. The values are as follows. Americas, 9, Europe, 5, Africa, 8, Asia, 38, and Oceania, 9.

Regions with countries and territories having VCs that cover landslide hazards around the world. Source: UNDRR Voluntary Commitments

In Asia, 38 landslide-related VCs cover a total of 15 countries involving 280 organizations. Data from various landslide databases show that Asia and the Americas are experiencing the highest risks and impacts from landslides (Gómez et al. 2023). The emphasis of VCs in Asia signifies a collective effort to address disaster risks and is a promising observation for landslide mitigation.

One such VC in the Asia region is Underlining the risk and preventing disasters for Atakum Municipality (ID 20230130_001). This is a local VC implemented by the Atakum Municipality local government in Türkiye. In line with the Provincial Disaster Risk Reduction Plan of the Samsun Province in the northern coast of Türkiye, Atakum Municipality is actively participating to fulfill its obligations and support the implementation of the local DRR plan. The municipality is dedicated to enhancing the community's disaster preparedness by implementing various measures, which include establishing a civil defense unit for disasters, creating designated disaster and emergency assembly areas, forming a search and rescue team, and utilizing a mobile earthquake simulation tool to increase the population's ability to respond effectively to disasters. This VC covers multiple hazards including earthquakes, floods, landslides, avalanches, and technical disasters. It focuses on supporting Priority 2 of the Sendai Framework, strengthening disaster risk governance to manage disaster risk, and Priority 4, enhancing disaster preparedness for disaster response. It also aims to contribute to target E of the Sendai Framework in the adoption and implementation of local DRR strategies in line with provincial and national strategies. Landslide risk mitigation and response is addressed here through the formation of a search and rescue team and the creation of disaster and emergency assembly areas.

An examination of several landslide databases show that China and Colombia are among the countries most significantly affected, along with India, Indonesia, the Philippines, Peru, Nepal, and Brazil (Gómez et al. 2023). In contrast, in the SFVC online platform, non-state actors from the Philippines and Nepal have submitted the most initiatives addressing landslides as presented in Figure 10. Although these countries were identified as among the most affected, there is still room for improvement and increasing landslide mitigation efforts by non-state actors in other high risk and high impact countries should be pursued.

Fig. 10
A stacked bar graph illustrates the number of V Cs covered by countries at regional and national or local levels. The Philippines has the highest value, with 9 V Cs at the national or local level and 2 at the regional level.

Countries covered by VCs (landslides). Source: UNDRR Voluntary Commitments

In the Philippines, there are 11 landslide-related VCs, two of which are regional while nine are national or local in scope. This involves 44 organizations with a commitment of 76 deliverables. One notable VC that addresses landslide risk in the Philippines is the Philippine Catastrophe Insurance Facility (ID_20221102_001), a private sector-led commitment with a national scope. It is implemented by the National Reinsurance Corporation of the Philippines, Philippine Insurers and Reinsurers Association, and the Philippines Insurance Commission. The facility aims to create a risk-based system for determining insurance rates that can sustainably cover disasters. The goal is to increase the country's financial resilience by supporting more people to access and afford insurance protection against catastrophes. The VC covers 16 out of 17 hazards that the SFVC online platform accounts for. If successful, the commitment would directly contribute to reducing economic less attributed to disasters (target C). The facility is also one of the few commitments on the platform that tackles insurance and risk transfer.

Organizations involved in landslide VCs are varied comprising 534 distinct organizations. There are 106 distinct implementers and 440 distinct partners. It is worth mentioning that organizations may participate in multiple commitments as either implementer or partner, thus the sum of the two does not amount to the total distinct organizations involved. Figure 11 shows the distribution of implementers and partners categorized into eight types with the lowest participation from news & media group and the regional organizations. Taking a closer look at the distribution of implementers and partners, we can observe that there is a much larger concentration of academic and research institutions as well as governments as partners in contrast with implementers. This could indicate that research institutions are inclined to be partners rather than direct implementers of commitments and this may be due to funding. It is crucial to point out that such institutions are a very important stakeholders for the implementation of the Sendai Framework as they can offer expertise, knowledge and capacity building especially in the themes of risk identification and assessment and science and technology.

Fig. 11
A double-bar graph presents the types of implementers and partners for V Cs. Non-governmental organizations have the highest value, while news and media have the lowest value.

Types of implementers and partners for VCs (landslides). Source: UNDRR Voluntary Commitments

The VC MYRIAD-EU: Multi-hazard and Systemic Framework for Enhancing Risk-Informed Management and Decision-making is a good example of a landslide VC that is implemented by a research institution, the Institute for Environmental Studies addressing the two themes for risk identification and assessment, and science and technology. The commitment involves a total of 16 partners from the private sector, other research institutions, networks, and non- governmental organizations. It covers 11 hazards including landslides and aims to develop a risk management framework that utilizes a comprehensive approach that considers multiple risks, sectors, and scales while also assessing trade-offs. Myriad-EU will be supporting five pilot areas to operationalize their framework through customized tools, products and services. This comprehensive framework is advantageous especially in the context of landslides, which are often triggered by floods and earthquakes, because it considers multi-hazard risk and impact assessments.

As previously mentioned, organisations can participate in more than one commitment as implementer or partner, and there are a number of organizations involved in multiple commitments (Fig. 12). In particular, eight organisations support four or more commitments. Remarkably, despite the low participation of UN & International organizations in the overall dataset, UN & International organizations make up half of the list, reflected in the figure as the last four organizations. The organization participating in the highest number of commitments is UNDRR as a partner to nine VCs.

Fig. 12
A stacked bar graph exhibits organizations involved in four or more landslide-related V Cs. The United Nations Office for Disaster Risk Reduction has the highest value for partners, and the CITYNET Yokohama Project Office has the highest value for implementers.

Organisations participating in four or more landslide-related VCs. Source: UNDRR Voluntary Commitments

Citynet Yokohama Project Office, which is considered under Network & Others, implements the largest number of VCs – two regional VCs encompassing Asia and two national VCs covering Nepal. This indicates that many initiatives are part of global partnerships and networks, emphasizing the importance of international cooperation in DRR. A further look into the participating organizations of each VC validates that many commitments involve multi-sectoral collaboration, including partnerships with local governments, private sector and NGOs

Finally, we also look deeper into the description section and deliverables section of landslide VCs through a quantitative and qualitative text analysis. The text analysis involved initial filtering to remove stop words, which are widely used words that provide little or no useful information. Sixty words were then identified for the description and the deliverables that accounted for the words with the highest frequency in each section. A second filtering was performed to remove “disaster”, “risk”, “resilience” and “DRR” as these are naturally terms that are often used to describe DRR initiatives. The end results show different ranking among the words with the highest occurrence and a few unique terms comparing the two sections. In landslide VC descriptions, the words with the highest frequency counts were community, local, climate, government and global. On the other hand, in the deliverables, the words with the highest frequency were city, business, government, local and community. The commonality between the two sections shows promising consistency in the focus on local, community, and government in landslide VCs. Additionally, it was also observed that more private sector related words were found within the deliverables such as business and MSMEs. We also discovered unique words that could be found only in the top sixty words in the description and unique words in the deliverables. From the filtered words within the deliverables, unique terms were found which pertained to specific outputs such as report, event, project, participants, forum, and network (Figs. 13 and 14).

Fig. 13
A word cloud for the description of the landslides V Cs. Some of the words are community, climate, local, government, development, Sendai, global, national, city, and building.

Word cloud for the description of VCs (landslides). Source: UNDRR Voluntary Commitments

Fig. 14
A word cloud for the deliverables of the landslides V Cs. Some of the words are city, business, community, development, government, continuity, international, national, and people.

Word cloud for the deliverables of VCs (landslides)

A qualitative review of both the description and deliverables section of landslide VCs show that common activities by non-state actors include capacity building and training, risk assessment and mapping of disaster-prone areas, development of emergency response and awareness campaigns. Some VCs show integration of technology such as artificial intelligence and unmanned aerial vehicles to conduct risk assessments and support early warning systems, while many initiatives emphasize research, innovation, and development of new tools and technologies for disaster risk reduction.

6 Conclusions

The analysis highlights that landslide VCs demonstrate a proactive and comprehensive approach to disaster risk reduction, encompassing a broader range of hazards, themes and issues, targets and indicators. Landslide VCs also involve a slightly higher number of partners. This indicates a collective effort to tackle landslide risks with long-term planning and a focus on engagement with various stakeholders. Landslide-related VCs have a more global and regional focus compared to all VCs. Geographically, a much larger concentration of commitments is focused on Asia, with the Philippines and Nepal implementing the most VCs addressing landslides.

Text analysis reveals a focus on local community, government involvement, climate change, global cooperation, private sector participation and urban cities. The commitments demonstrate a diverse range of activities and approaches to DRR, reflecting the need for multi- sectoral collaboration, technology integration, and community engagement in building resilience against disasters.

It should be emphasized that many initiatives by non- state actors have yet to be reflected in the platform. To conduct a more robust analysis, more stakeholders should be encouraged to submit their commitments to the platform for deeper insights into their contribution to disaster risk reduction with focus on different hazards such as landslides. Additionally, for future studies, it may be useful to conduct a survey among active stakeholders about interventions within their commitments specific to landslide risk reduction and response.

The prevalence of landslides globally underscores the need for continued efforts by non-state actors encouraging initiatives in particularly vulnerable countries. Strengthening partnerships with research institutions and the private sector could foster innovative and effective solutions for landslide risk reduction and address the need for broadening risk assessments and data sharing.