Blog posts

ForPAc has been working to improve early warning systems and anticipatory preparedness for flood and droughts. Building on existing established relationships between UK and Kenya, the team first established a participatory process to bring together the agencies mandated for provision of forecasts (KMD, ICPAC and RCMRD) with risk management, response and preparedness (NDMA, County Govt ministries and KRCS).

This process mapped out the complex existing EWS systems covering both drought in mostly rural, arid regions, and flood risk for fluvial and pluvial flooding in both rural and urban contexts. The mapping covered information flows, response interventions, responsibilities and resource and finance issues.

The project research then directly addressed the two key challenges: improving forecasts and improving preparedness actions.

Forecasting was strengthened by drawing on the best available international science and global and regional modelling and evaluating forecasts of flood and drought metrics. The team then developed a suite of new forecast products for useful indicators like rainfall, vegetation, and soil moisture for drought, and river flow and inundation maps for floods, across a range of forecast lead times, from days to months.

'ForPAc has improved the capacity of KMD to take advantage of the major advances in global forecasting and the appetite of our national risk management agencies, to respond to this opportunity.' — Director, KMD.

'[RCMRD] now use the [ForPAc] algorithms...and software...into our own pipeline [to] produce the novel VCI forecast product, [which is] qualitatively different from anything we've been able to do before and provides our stakeholders with a decision-making tool with much greater power.' — Director General, RCMRD.

Preparedness actions were strengthened by designing anticipatory decision-making processes, within the existing systems, triggered by these forecast products to mitigate hazard impacts. These new forecast products were piloted with the relevant risk management agencies, building understanding of the products and their use in evidence-based decision-making. The Kenya NDMA is now incorporating drought forecasts in their EWS bulletins with advisory actions, with the intention to develop a forecast-based drought EWS (rather than the existing reactive system) across all the arid and semiarid counties of Kenya, the Kenya Red Cross Society has incorporated the drought and flood forecast into the new operational early action protocols.

'Following [ForPAc] research, several novel and skilful forecasts of key drought indicators... have been co-developed and piloted in several counties. As a consequence, the NDMA has developed a new template for its monthly drought bulletins...The template now includes...[these] forecasts...in order to reduce the impact, recovery time and costs associated with traditional drought response...giving stakeholders ample time to initiate drought preparedness actions…These projects have substantially advanced our drought early warning system...[ ForPAc] built trust of the decision makers in these forecasts and the capacity of stakeholders in the counties to interpret... forecast... and probabilistic information has greatly improved.” — CEO, NDMA.

'[ForPAc] supported decision making at KRCS [leading to] concrete changes in its practices in disaster risk management and underpinning a paradigm shift towards a more anticipatory approach.' — Head of research and training, KRCS.

'[ForPAc activities represent] a template of how to move towards anticipatory risk management within existing national frameworks…“[T]he ForPAc project has supported and advanced the technical aspects of Forecast based Finance application, influenced the way FbF is being scaled up by highlighting the importance of mainstreaming, while underpinning capacity and dialogue between national Red Cross societies, risk management agencies and the global FbF movement...[ The research provided] the technical capacity to guide the development of [anticipatory action] in Africa.' — Director, Red Cross Climate Centre.

Despite the range of PhD focus areas, some commonalities emerged from the reflections, including the need to focus on user requirements to drive innovative approaches to new data and improved forecasting science, and the benefits of working closely with stakeholders and communities to co-develop appropriate information that can be used in real-world decision making.

The students will be building on from this virtual workshop to discuss collective lessons to share via knowledge products and during the SHEAR finale event in 2022.

Find out more about each PhD topic.

Some recent publications shared by the PhD students include:

• Gurung, L, and McGowran, P. 2021. Local perspectives on landslide prevention and management in Kalimpong district, West Bengal, India. Disaster Prevention and Management. DOI: https://doi.org/10.1108/DPM-02-2021-0056
• McGowran, P, and Donovan, A. 2021. Assemblage theory and disaster risk management. Progress in Human Geography, Vol. 45(6), 1601–1624. DOI: https://doi.org/10.1177%2F03091325211003328
• Paul, J, Buytaert, W, and Sah, N. 2020. A technical evaluation of Lidar-based measurement of river water levels. Water Resources Research, Vol. 56(4). DOI: https://doi.org/10.1029/2019WR026810
• Russell, C, Clark, J, Hannah, D, and Sugden, F. 2021. Towards a collaborative governance regime for disaster risk reduction: exploring scalar narratives of institutional change in Nepal. Applied Geography, Vol. 134, 102516. DOI: https://doi.org/10.1016/j.apgeog.2021.102516

We are pleased to share the collective outputs from the workshop:

• a publication on the discussion points and findings, including examples from SHEAR projects
• reflections from the workshop (video)
• a journal article related to the workshop activities
• ten individual video presentations used to demonstrate examples of where mobile phone technologies have been used in DRM

Opportunities

Increased mobile phone ownership in developing countries can support positive societal intervention because they provide a way of connecting directly and immediately to more people. Some of the opportunities of using mobile phone technologies for DRM are:

• instant data: mobile phone technologies have the potential to provide access to instant and immediate data, which aids early response action
• widespread communication: mobile phone technologies can enable widespread and real-time communication bewteen experts and vulnerable communities
• two-way communication: there is opportunity for two (or more) way conversations through mobile phone technologies where communication between citizens, scientists and disaster managers is possible
• education and empowerment: mobile phone technologies can provide a way to educate and empower people by increasing access to hazard-related information and knowledge

Lessons

Despite the opportunities mobile phones create, using them for DRM can present some challenges. Some of the considerations that may help to improve either safety, efficiency or effectiveness of mobile phone technologies in DRM were discussed during the workshop.

• Avoid parachute solutions: local dynamics, contexts and needs should be understood before utilising or creating new mobile apps. Ideally, mobile products should be demand driven, involving action-based training and a human-centred design methodology to elicit user requirements. This involves avoiding 'parachute solutions' and instead encouraging research that understands the needs of specific communities before before developing any solutions.
• Recognise vulnerable community members: vulnerable members of communities, such as the elderly and women, should be recognised within design. Design that recognises existing socio-economic vulnerabilities allows for more equitable access to information and equitable participation in data generation.
• Language and literacy: factors such as language and literacy should be discussed before apps are developed, to ensure relevance to target communites.
• Data ownership and protection: having a clear understanding of the intellectual property of the app or system and a clear long-term plan for how data will be appropriately managed and maintained is important to think through from the outset. Having clear and transparent data management protocols, especially when social media platforms such as WhatsApp and Twitter are being used, is imperative for trust and success.
• An app is not a panacea: recognise that an app is not a panacea and may not be the most appropriate solution for the users in certain contexts. For example, the Landslide-EVO project demonstrated that some elderly populations in Nepal did not own smartphones and were not interested in eganging with them anyway.
• Internet access: variability in internet access, due to cost and connectivity, was outlined as potentially prohibitive for some communities that were using mobiles in disaster risk reduction.

Main take-aways

The SHEAR research presented during the workshop demonstrated that mobile phones have promising potential for all elements of DRM. Mobile phone technologies have the potential to provide access to instant and immediate data; this increases rapid and reliable dissemination of information and action during a disaster in the short term, educates and empowers local people in the medium term, and ensures scientific capacity building in longer term.

The workshop highlighted that effective use of mobiles in DRM requires appropriate consideration of local context and should continuously and iteratively adapt to cater to the needs of the users and stakeholders; and, ultimately, project design should be human-centred and demand-driven as far as possible.

During the MayDay Round-the-clock event, SHEAR's LANDSLIP consortium members shared their knowledge and experience via presentations and coordinated sessions on a variety of key topics.

• Christian Arnhardt (British Geological Survey) co-chaired a discussion on national and local landslide early warning systems (LEWS). In this session, operators and managers of landside early warning systems discussed similarities and differences in the organisation of national and local LEWS and how to promote their interaction.
• Joanne Robbins (UK Met Office) co-chaired the communicating and engaging with stakeholders working group session, presenting findings from a recent group questionnaire and identifying key challenges that might direct next steps. There were also discussions on key topics, such as:
  • practical approaches to engagement (how often, style, terminology, language, focus, and scope)
  • roles and responsibilities of different actors within LEWS
  • ensuring effective pull-through and longevity of researcher-user interactions
• During the session on methods, opportunities and challenges of landslide data, Alessandro Mondini (Consiglio Nazionale delle Ricerche) presented SAR for landslide failures detection and mapping: use and perspectives.
• Saibal Ghosh and Akshaya Mishra (Geological Survey India) participated in a panel on perspectives on innovations from LEWS operators, discussing how they deal with new ideas and methods that are promising but not yet established. The panel discussed strategies for implementing innovations and major needs, barriers and chances with regard to the implementation of innovations.
• Mirianna Budimir (Practical Action Consulting UK) shared existing lessons and best practice on stakeholder engagement and communication from the wider EWS field that can be applied specifically to LEWS, using examples from operational LEWS and the SHEAR programme. The session also invited sharing of experiences, challenges and strategies from the LandAware community.
• See the full video recordings.

However, gaps remain between the production and use of forecast information, meaning that information can be produced which is not fit for purpose, and which excludes the experience and expertise of directly affected people and the institutions which seek to support them, including national, local and regional governmental and non-governmental decision makers.

Co-production is an approach that aims to bridge these gaps by bringing together the different stakeholders to work collaboratively to produce forecast information which is relevant, accessible, and useful for decision makers across scales.

Putting knowledge into practice

Learning to Co-Produce (L2CP) is a SHEAR initiative which has been developing an online training to strengthen meteorologists' capacities to develop weather and climate information that can better support the needs of decision makers. The training is oriented towards early career producers of forecast information and aims to identify the key steps, skills and processes for producing user-centred weather and climate information.

The L2CP training has been co-developed with a range of meteorological training and climate research institutions across East and West Africa and the UK. Co-ordinated by King's College London, partners include:

• University of Nairobi
• Agence Nationale de l'Aviation Civile et de la Météorologie (ANACIM)
• Université Cheikh Anta Diop (UCAD)
• Kenya Meteorological Department (KMD)'s Institute of Meteorological Training and Research (IMTR)
• Uganda National Meteorological Agency (UNMA)
• IGAD Climate Predictions and Application Centre (ICPAC)
• Kenya Red Cross Society (KRCS)
• Walker Academy at the University of Reading
• universities of Bristol, Oxford and Sussex
• UK Met Office
• UK Centre for Ecology & Hydrology (UKCEH)

'Before I did not know that the public are also important in climate issues, in my research I'll be engaging the public to get more about climate and how it affects them.' — L2CP learner

Designed to strengthen skills in engaging with forecast users, the course draws on experience of developing weather and climate services for different countries and contexts across West and East Africa. Modules explore topics including the principles of co-production, key climate concepts and conveying these to non-technical decision makers, understanding decision-making contexts, and monitoring and evaluation. Combining theory and practice, the course is designed to support remote and flexible learning.

With the aim of supporting its integration within meteorological training curriculum, the course was piloted between October and December 2020 with 105 students and staff from across 11 institutions. Most learners assessed that the course had considerably strengthened both their understanding of what is entailed in the process of co-producing weather and climate services as well as their confidence to engage in this process.

'The course was truly a wonderful experience. It has helped me appreciate not just the role of co-production, but also boosted my own confidence as a meteorologist. It has endowed what I do with meaning, by showing me that there is a way to make the work we do more worthwhile to the development endeavours… Being a meteorologist is now endowed with greater meaning, direction and purpose.' — L2CP learner

What next?

L2CP has undertaken an evaluation of the course's initial piloting, seeking students' and partnering institutions' feedback on how to further strengthen the course. Review has highlighted the need to:

• provide more time to complete assignments
• seek additional support to enable translation of the course into French
• bring in more practical examples and demonstrations of key concepts
• increase opportunities for interaction among students and with course tutors

L2CP reflects and implements SHEAR's goal of working with stakeholders to co-produce demand-led, people-centred science and solutions. Following this pilot, three institutional partners — the Kenya Meteorological Department, the University of Nairobi and the Ugandan National Meteorological Training School — have proposed using elements of L2CP in their future training. The L2CP team is also exploring potential to work with institutions providing climate training in Senegal and the African SWIFT project, as well as sharing the resource via the WMO Global Campus. The team have also been invited to present the course at an ICPAC (IGAD Climate Prediction and Application Centre) learning event in March 2021.

'With the current changes in weather and climate, co-production is definitely the only way to go. Weather and climate impacts all sectors of the economy and therefore, to reduce the impacts, we need to start taking in this weather and climate information. However we can only take it in if it is useful to us (tailored to our needs) which necessitates us to engage in the production of this weather and climate information for easy understanding and interpretation of the forecasts.' — L2CP learner

By focusing on bridging the gaps between research and practice, this course is working to strengthen effective preparedness for increasing climate extremes, variability and change.

• Full report of the evaluation of the pilot course
• To find out more about the Learning to Co-Produce project, please contact Emma Visman

NFLICS

Nowcasting FLood Impacts of Convective storms in the Sahel

Flash flooding and intense rainfall in parts of Africa cause loss of life and damage of food and property. In the SAHEL region, intense rainfall is caused by Mesoscale Convective Systems (a complex of thunderstorms that becomes organized on a scale larger than the individual thunderstorms that normally persist for several hours or more) during the rainy season. NFLICS uses historical weather data from satellites from 2004-2019 to develop automated nowcasting i.e. weather forecasting on a timescale of around two to six hours. This reduces the need to maintain consistent on-ground observational weather networks and creates a large data set that can be used for forecasting future weather patterns.

FATHUM

Forecasts for AnTicipatory HUManitarian action (FATHUM)

Part of the FATHUM research is to look at improving the lead-time of forecasts of river flooding in sub-Saharan Africa. Research into the role of climate variability in driving flood risk has shown that the Indian Ocean Dipole has just as much influence as the El Nino Southern Oscillation, and these modes of climate variability have a dramatic impact on the timing of the flood wave, as well as the magnitude of flooding. Research into how lead-time influences the climatology of medium-range global flood forecasting systems has been put into practice and used operationally within Red Cross Early Action Protocols, improving the skill of the forecasts at longer lead-times.

INPAIS

There are limited observations of typhoon events which makes it difficult to have a reliable typhoon risk assessment, especially for the use in parametric insurance solutions. INPAS aims to address this by constructing a physically consistent typhoon event set based on unrealised typhoon events in operational forecast outputs. These typhoon events follow known physical laws and the amount of events in this event set is comparable to current reinsurance industry standards, but can better include events in locations and of strength not observed yet. Parametric insurance offers pre-specified financial pay-out based upon a trigger event. This means that compensation or relief funds can be quickly distributed to communities most vulnerable to typhoons.

LANDSLIP

LANDSLIP

Forecasting rainfall induced landslides is difficult because of spatial and temporal uncertainties in rainfall forecasts in mountainous regions. Landslide forecasting has been dependent on nowcasting data that predicts 1 to 2 days ahead which puts pressure on landslide warning communication and dissemination in India. LANDSLIP’s weather forecasting prototype has the potential to identify local level heavy rainfall and landslide patterns up to 15 days ahead, which allows experts users enough time to prepare and alert response agencies with a greater lead time.

NIMFRU

National-scale IMpact-based Forecasting flood Risk in Uganda

The NIMFRU project is engaging with researchers, policy makers, local councils and local farmers to improve the targeting, relevance and communications of flood warning and response in Uganda. The project has been collating and building the climate information data sets for the Katakwi district in Uganda and has integrated data from the IDAPs platform, the Rain watch Platform, pre-existing data sets from the National Disaster Agency NECOC and Glofas modelling data. These data have also been integrated with new Household Economy Approach data sets collected from the district. The integration of these interdisciplinary data sets have resulted in the Livelihoods Impact-Based Forecasting (LIMB) system, which has been built to support policy makers with improved forecasting and their decision making processes for flood response.

What have SHEAR projects achieved in relation to improving data?

These projects are examples of SHEAR projects that focus on improving data generation.

BOULDER

BOULDER

BOULDER focuses on developing early warning systems for landslide flood hazard disaster evaluation in the Upper Bhote Koshi catchment in Nepal, one of the most landslide and flood prone countries in the world. BOULDER focuses on the development of low-cost technology to build sensors which can be attached to boulders and sense when they start moving. This project addresses specific landslide and flood risk-management problems brought up by community stakeholders impacted by boulders.

WeACT

WeACT

WeACT uses high-performance dam-break flood modelling to innovate natural dam-burst floods (NDBF) hazard assessment and forecasting systems to improve community flood preparedness and resilience in Nepal. Catastrophic floods resulting from the failure of dams that impound glacier lakes are known as glacial lake outburst floods (GLOFs). GLOFs and NDBFs have been recognised as one of the most serious natural hazards in populated mountainous regions across the globe.

HYFlood

HYFlood

Flood hazard and risk maps create a database that cover issues such as land-use planning, insurance and capital provision, emergency response and disaster preparedness. However, across most of sub-Saharan Africa such data is absent which poses a huge challenge to disaster risk managers. HYFLOOD improves understandings of the occurrence, location and intensity of flooding by building on an existing global flood model to develop regional to community scale flood hazard maps in sub-Saharan Africa.

Landslide-EVO

Citizen science for landslide risk reduction and disaster resilience building in mountain regions

Landslide EVO has looked into citizen science and participatory approaches to co-generate locally relevant actionable knowledge on disaster risk reduction. Among other things, the project combines community-based environmental sensing with satellite data to generate maps and forecast landslide triggers to map and model landslide risk and vulnerability along the Seti River in the Karnali Basin, Nepal.

NIMFRU

It has improved the data used by national agencies for flood response by adding and integrating livelihoods data collected from the Katakwi district to existing and new modelling data of the region. This household livelihoods level data has been collected using the Household Analysis Approach, and provides essential ground level data and understanding on different coping thresholds and vulnerabilities of individuals to ensure that flood responses can be targeted to those who need them the most.

What new tools have SHEAR projects developed?

SHEAR projects have been involved in the creation of cutting-edge tools and technology. These technologies include:

• ForPAC, which works to improve forecasts for flooding and drought in Kenya so that decision makers are able to take early action to reduce the impacts of these hazard events.
• COSMA, which aims to develop an integrated modelling approach by taking into account the urban heat island, building characteristics and vulnerable population to build effective early-warning systems and a city-scale heat action plan in Sri Lanka.
• C-FLOOD, which is producing a new generation of compound flood-hazard maps, based upon state-of-the-art computer modelling that will consider all the storm-surge components and the rainfall effect in Sri Lanka.
• IPACE-Malawi, which investigates the impacts of extreme weather events on agricultural systems and contribute to improving the forecasting and delivery of agriculture-specific weather information to improve preparedness of farmers and humanitarian and disaster response organisations.
• NIMFRU, which integrates the essential livelihoods data into flood response, has built a multi-sectoral Livelihoods Impact-Based Forecasting (LIMB) system which is a piece of free access software that policy makers can use to access the livelihoods and modelling data to inform their decision making.

Overall, SHEAR projects have demonstrated tangible efforts in improving forecasting science by testing new approaches and assessing how well they perform for application purposes. The projects described in this article are examples from a broader set of SHEAR initiatives that can be explored further on our website.

Watch our video exploring these new developments.

SHEAR at the Understanding Risk forum

The SHEAR session on landslide early warning presented an overview of such systems from both a technical and operational/practical perspective, drawing on experiences and knowledge across the globe and case studies of India and Nepal by the SHEAR LANDSLIP and Landslide-Evo projects. Presenters and panelists included representatives from Practical Action, the British Geological Survey, King's College London, Imperial College London, Consiglio Nazionale delle Ricerche, the UK Met Office, the Geological Survey of India and the Geological Survey of Austria. Ten individual presentations were recorded and are showcased for public access.

The keynote session presented a high-level overview of landslide early warning systems and also highlighted the importance of the physical, social, and practical trifecta in building and developing these systems. Five sessions then focused on the physical science aspect of these systems.

• the components and functionality of regional landslide forecasting
• the mapping of landslide susceptibility
• using weather patterns for landslide forecasting
• data needs for landslide forecasting
• low-cost monitoring local landslide early warning systems: a case study of Western Nepal

Four more sessions focused on the social science aspects of these early warning systems.

• institutional mapping and stakeholder engagement to ensure the sustainability of these systems
• citizen science and participatory monitoring
• developing informational tools for decision makers, in the form of bulletins, early action protocols and standard operating procedures
• reflections on the development of a prototype regional landslide early warning system in India through the SHEAR programme

You can now view the entire session and we encourage you to get in touch with us if you are interested to know more about this work.

You can sign up for SHEAR news.

Addressing this challenge

The LANDSLIP project has been working to address this challenge by exploring and testing an approach which greatly simplifies the complex process of preparing and interpreting the maps, reducing the time and needed to train the users of the maps, and subsequently for the maps to be produced and used.

The approach is based on a binary classification of each pixel in an aerial image of a land area, resulting in a simplified land cover map which specifies only whether an area is 'landslide' or 'no landslide'. Compared with images which include numerous land cover categories (such as vegetation, urban areas, water bodies, and landslides),

This approach combines landslide maps obtained from satellite imagery with supporting spatial information on the other factors facilitating their occurrence, to reduce the number of false positives (or false alarms).

By improving the performance and the cost- and time-efficiency of this vital process for geological agencies and decision makers in the fields of preparedness and response, this work can support efforts to build resilience to a hazard which causes such devastating impacts for communities.

Questions and areas requiring further research

In this brief, the authors highlight two main questions, each with respective subsidiary questions.

What are the design and operational factors that enable or constrain effective scaling up of forecast-based anticipatory action through social protection?

This question highlights the high urgency for evidence-building prior to establishing linkages. Indeed, limited understanding about the potential and relevance of linking national social protection systems and forecast-based action has been identified as a primary barrier to the development of such mechanisms. Addressing this question can have high potential to promote successful integration in the future, as identification of these design-related and operational factors is the first step in filling the gaps and addressing the challenges.

Which political, institutional, and financial factors are key to enabling further integration of social protection?

This question underlines the importance of factors other than programme design and operation that can bolster (or hinder) the integration of social protection and forecast-based action. Again, addressing this question has significant potential in successful integration at scale.

The concluding output from the research is a checklist to inform future scoping studies, identifying pressure or entry points for anticipatory social protection, and could be used as a starting point for project teams to explore the scope of this research area.

Three major conclusions come out of this brief. First, investing in anticipatory social protection is important and effective, and robust forecast-based action can complement the development of such systems if and whenattention is adequately paid to complete vulnerability assessments, database management, and targeting design. Second, clear precedents and MOUs must be set, in context-specific approaches that takes vulnerability and political economy into account before integrating and scaling social protection and forecast-based action. Third, a proper understanding of financial capacity and funding sources is necessary to realistically harness the potential of integrating social protection and forecast-based action.

In answering that question, I make a couple of unrealistic assumptions. I assumed that I could have travelled to my field site and that I could have stuck to a similar methodology. In reality, I wouldn't consider travelling and speaking to lots of different people, as this would put myself and others at risk. Nonetheless, I think it is an interesting way to reflect on the research process and the way in which it is produced, not just by the researcher but also by the people and processes the researcher works with and around during fieldwork. This idea relates to a number of theories in geography and science and technology studies (STS) relating to the coproduction of knowledge, and specifically the role of 'the non-human' in this process (Donovan, 2017).

Fitting my research in and around the monsoon: reality 2019

I spent a total of six months in my field site of Kalimpong District, West Bengal, India. This was split between two trips, one visit of two months between May and July 2019 and a second, more intensive, visit from September 2019 until January 2020. The purpose of the first visit was to build a network, get a feel for the area and experience some of the monsoon season and associated landslides. I returned in September with a refined methodology, a network to support me through the data collection phase and some experience of the monsoon season and landslides.

What if...2020

This may have not been how things would have panned out had I done the fieldwork in 2020. Forgetting about university policies that would have prevented me from travelling for research, I may have chosen to just do one longer trip to prevent the need to travel back and forth. All of this may have affected the amount of data I collected, who worked with me and the issues discussed. So, for the remainder of the blog, I will assume that Covid-19 would have made me complete one six-month field trip from June to December 2020. I believe two major differences that would have affected my research would have been the heavier monsoon and the existence of coronavirus.

First visit (2019) — where is the rain!?

After conversations with the leader of a local NGO who works on disaster-related issues, I came to the conclusion that for a seven-week stay in The Hills from the end of May until mid-July, I should experience at least a few weeks of the monsoon in Kalimpong. Unfortunately — for me and many people relying on agriculture in Kalimpong — June rainfall was extremely deficient, at a mere 81 mm! This was considerably less than the also relatively dry Junes of 2017 and 2018, totalling 154 mm and 184 mm respectively. Had I visited in June 2020, I would definitely have made more use of the umbrella I bought at the start of my trip! Save The Hills reported eight times as much rainfall in Kalimpong in June 2020 than in June 2019, with a considerable 651 mm falling in June alone. The heavy rain continued through the monsoon of 2020, as follows:

• July 751mm
• August (634mm)
• September (580mm)

(See Save The Hills for more data.)

In July 2019 the values were:

• July (821mm)
• August (510mm)
• September (408mm)

Comparing the two years' values it is clear that there would be more landslides in Kalimpong in 2020 than in 2019. In reality, the only serious monsoon rain I experienced occurred the week I left Kalimpong (see my blog post about this).

Second visit 2019 — reflecting on the temporalities of my data

During my second visit, we completed 62 recorded interviews and one focus group discussion, speaking (on record) to around 100 interviewees (interviews were often with more than one person). Of these, 45 interviews were with people affected by landslides, with the other 15 with 'key informants (academics, NGO workers, journalists, officials etc). Most of the interviews were done with my research assistant Lochan, who I befriended in a local café and community hub during my second visit. He acted as a translator, driver, interview organiser and source of important knowledge on local customs and news. I could not have done the research without him. As part of the interview process, people would often show us around the area around their home or where the landslides had occurred. This, combined with innumerable informal conversations, extensive field notes and simply being in The Hills for an extended period, was as vital to the research process as the interviews.

One of the reasons for not being in The Hills for the whole of the 2019 monsoon was the access issues this would have caused. However, this may have been mitigated by an extended period of fieldwork in 2020, where I could have done research in spurts during drier periods. This would have also allowed me to appreciate the rhythms of life in The Hills during the monsoon more clearly. Conducting interviews during the heavy monsoon of 2020 and consequent increased rate of landslides may have also made descriptions of landslides 'as they happened' that people relayed to us in interviews 'thicker'. In reality, we struggled to speak to people affected by recent landslides in Kalimpong District, largely because very few had occurred since 2017, with most of the ones referred to by my interviewees occurring in 2015. This time lag affected the type of information gathered during interviews, as the data contains multiple stories about the long-term impacts of landslides on people's lives. Whilst this is important, a combination of different temporalities of landslide impacts may have been of benefit to my dataset and research. Overall, the type of data that I have and thus the thesis I will write has been shaped significantly by the monsoon conditions of the year that I visited.

How would Covid-19 have changed my research further?

By considering some research that has been done on the impacts of Covid-19 in The Hills (Kharang et al., 2020), and the data from my interviews, it seems clear that one of the most important and devastating connections between Covid-19 and landslides is the impact on food security. We found that landslides profoundly affect food security, most obviously through destroying or removing the land that many people in Kalimpong live off and through diverting water sources and rendering the land infertile through soil erosion. The impacts of these processes on livelihoods are felt in the long term and cascade through generations (see more).

ATREE, an NGO that has a branch in Darjeeling, completed a rapid assessment of the impact of Covid-19 and the associated lockdown in The Hills (Kharang et al., 2020). Their findings suggest a considerable impact on food security. This was caused by factors such as the closure of markets that sell imported food from the plains, the decreased ability to farm their own land due to the lockdown coinciding with sowing season and a decrease in tourist numbers. This, combined with the longer term degradation of food security due to landslides, climatic changes and crop viruses that people told us about in interviews, points to the need for long-term policy intervention to support local food production (Kharang et al., 2020).

The combined impact of a heavier monsoon and Covid-19 would have led to quite a different dataset and thus research outputs from my PhD. In many ways, Covid-19 will still affect my research and will have to be considered in my thesis. Fortuitously, I am able to draw upon research done by people who are living in The Hills and experiencing and reporting these issues (Kharang et al., 2020).

Overall reflections — can we make disaster research more resilient?

Combining this with the fact that the research I did was only possible with the support of a local research assistant, there is another conclusion to be drawn. That is that disaster research might be more resilient to changes in the material world if local people were empowered to conduct and lead more of this type of research in the places where they live. This dovetails well with a number of arguments emerging from the literature relating to the need for knowledge to be co-produced (Gaillard and Mercer, 2013). Whilst I have toyed with the idea of being able to conduct my research during these times of Covid-19 for the benefit of writing this post, the reality is I wouldn't have been able to do it — the ethical and logistical issues are just too large.

In fact, I don't know what I would have done and I sympathise immensely with PhD students who are having to grapple with changing their methodologies. So, whilst I am not suggesting that PhD researchers shouldn't be able to travel and conduct this kind of research, it is difficult to argue against the fact that if more local researchers, as well as PhD researchers, were conducting the type of research I was given the opportunity to do last year, with the same amount of funding and suppor, they would actually be able to adapt their methodology to the conditions and be able to continue data collection. This would allow them to capture the combined impacts of Covid-19 and a heavy monsoon season on the people of Kalimpong.

This would work towards addressing the specific issue in Kalimpong/Darjeeling of young people feeling as though they need to leave The Hills in order to pursue these kinds of opportunities, which was an issue for many of the people I spoke to in Kalimpong, who I now call friends, would frequently mention. This more locally led research would also be more likely to bring about long-term change in those places (Gaillard, 2019), which would benefit the people of Kalimpong who are affected by disasters. In turn, this would be of benefit to the study and practice of disaster risk reduction (DRR) more generally.

References

Donovan, A. 2017. Geopower: reflections on the critical geography of disasters. Progress in Human Geography, Vol. 41(1), 44–67.

Gaillard, J C. 2019. Disaster studies inside out. Disasters, Vol. 43(S1), S7–S17.

Gaillard, J C, and Mercer, J. 2013. From knowledge to action: bridging gaps in disaster risk reduction. Progress in Human Geography, Vol. 37(1), 93–114.

Kharang, S, George, R, Darji Bhutia, T, Pradhan, S, and Pradhan, A. 2020. Rapid assessment of the impacts of Covid-19 on the livelihoods of rural communities in Darjeeling and Kalimpong Districts, West Bengal.

Accessible, high-quality training despite pandemic

The Covid-19 pandemic ruled out plans to deliver face-to-face training courses on the results of new analysis of recent cyclones in the Southwest Indian Ocean (the PICSEA project) by Dr Rebecca Emerton, Dr Nick Klingaman, Dr Kevin Hodges, Prof Pier Luigi Vidale of the National Centre for Atmospheric Science at the University of Reading and Dr Liz Stephens of the University of Reading. Instead the IEA worked with them to develop a multimedia online training course for operational forecasters in Mozambique, Madagascar, Seychelles and at the Red Cross Red Crescent Climate Centre.

From socially-distanced video interviews, to appealing animations and a fascinating podcast with four world-leading experts, learning activities were created from the fundamental research and collated into one website for simple and engaging access. The variety of formats and tailoring have been selected with the learners in mind and it includes a quiz for self-assessment. The activities are mostly brief, so busy forecasters working shifts can dip in and out of the materials as time allows, with the online course continually available. Content is provided in English, Portuguese and French.

Weather research at the University of Reading

Dr Emerton studied the tropical cyclones of the south-west Indian Ocean of the last decade, investigating how operational computer modelling had advanced over that time. The results of this work provide useful guidance to the forecasters regularly interpreting model outputs, giving a better understanding of the uncertainty in the projected track of a cyclone and the likely impacts of severe winds and rainfall. The studies included the effects of the Madden-Julian Oscillation — areas of enhanced and suppressed tropical rainfall which can encourage or discourage the development of tropical cyclones.

When Cyclone Idai caused catastrophic floods and landslides in Mozambique in March 2019, followed just weeks later by Cyclone Kenneth, Dr Emerton was able to apply the research to provide real-time flood warnings to help local authorities and humanitarian organisations make informed decisions on deploying aid to those most in need.

Dr Klingaman said: 'Tropical cyclones are a serious threat to south-east Africa, with impacts regularly experienced in this little-studied region. This training will contribute to forecasters’ understanding of these devastating events and to which communities they should be sending warnings. Humanitarian organisations will be better informed about where to target anticipatory action.'

Training for weather forecasters worldwide

Colin McKinnon, CEO of the IEA, said: 'This is another example of the IEA working effortlessly across the world, using our expertise to deliver tailored insights and capacity-building on weather and climate, contributing to regional resilience.'

The training for weather forecasters is a free and open resource for all, appealing to tropical meteorologists and engaged forecast users, especially those dealing with the impacts of tropical cyclones.

• Training page: https://www.the-iea.org/picsea/

The training course was commissioned by PICSEA, which is funded by UKRI SHEAR and the research was carried out by the National Centre for Atmospheric Science (NCAS) at the University of Reading. The PICSEA partners are the national meteorological centres of Mozambique, Madagascar and Seychelles and the Red Cross Red Crescent Climate Centre and was support by the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Met Office.

To find out how the IEA can create and deliver bespoke training for your products email Vicky Lucas.

Recently, citizen science has emerged as a useful tool in WRM to energise local stakeholders, placing them at the heart of decisions regarding water resources (Paul et al., 2018; Paul and Buytaert, 2018). The data that citizen scientists collect can strengthen traditional datasets in data-scarce regions and the rapid technological advances that have made devices like smartphones newly accessible mean that participatory data collection is more feasible at a more sophisticated level than previously.

Installing low-cost river-level sensors and rain gauges that citizen scientists can read, downloading data which can be transformed into actionable information and shared with community members using smartphone apps or SMS alerts, for example, presents an example of improving observational networks in this way.

These kinds of participatory initiatives can also support communities in increasing engagement with and understanding of scientific concepts, approaches and methods and can support communities to be empowered as decision makers and stakeholders in governance systems and structures. For participatory citizen science and engagement to be effective, it needs to be approached in a way that is based on a meaningful understanding of the context and is sensitive to the different perspectives, experiences, needs, priorities and vulnerabilities of the range of stakeholders involved. There is a possibility of new scientific insights causing apprehension in a community, while poor or inappropriate communication of results could lead to misunderstandings with long-term implications and effects.

Engaging in regular mentoring and developing feedback loops are ways to build trust, hared learning and understanding between scientists and citizen scientists, which is vital for the sustainability of a project’s impact.

You can find out more about the Landslide-Evo project and you can read the full paper 'Improving water resources management using participatory monitoring in a remote mountainous region of Nepal' to find out more about the project’s approach to and learning about citizen science.

This information is important for the Forecast Based Financing (FbF) & Forecast Based Actions (FbA) especially in cautioning our vulnerable pastoralist at a broad scale. However, to adequately address such challenges at micro–scale, would ideally require a similar level of understanding in vegetation conditions. For example, livestock and forage insurance have been identified as a resilience measure which would ideally mean understanding micro–scale vegetation dynamics.

What does a stationary camera infer?

It’s important to understand the growth cycle of plants if we are to detect these disturbances (such as precipitation). It's equally important to have information on species composition in land cover if we are to have an effective resilience measure such as one mentioned in the previous section.

Stationary cameras (PhenoCams) provide repeated images acquired in a time series with 30min intervals, which is crucial in helping us understand how ecosystem processes are affected by such disturbances in season.

Using computed greenness index calculated from standard red, green and blue colour stored in each image, precise identification of changes in vegetation can be detected. Additionally, PhenoCams provide a species–level understanding of such vegetation which is deemed important for not only understanding the condition of vegetation but also to identify the vegetation species (forage) which is required for the livestock.

From figure 2, its clear different vegetation species respond differently to disturbances thereby making detection of drought episodes possible. Additionally, sustained pastoralism hinges on the palatability of the forage material, hence PhenoCam provides a detailed view of these dynamics by providing detailed in–situ high–resolution measurements. Such information is deemed important as it provides an up–close & detailed understanding of vegetation dynamics. A network of similar systems set up in ASAL areas provides localised vegetation conditions as well as a platform for validation of the space–borne platforms.

Conclusion

To effectively address the risk posed by drought requires a good early warning system in which decisions are made based on accurate information. Resilience and adaptation of such risks require complex integrated systems. Thus, the methods developed in this research helps to identify deteriorating vegetation, well enough in advance to be of real benefit in helping mitigate the impact of drought in ASAL regions.

Taming the wild river

In 1869 and 1870, floods in the Kosi River resulted in significant damage to crops in India, and the river's unpredictable shift was regarded as a threat to the planned development activities in the Kosi River basin. In 1893, a team of British engineers studied the Kosi River to identify possible flood protection solutions and produced a report for the British government. The report rejected the possibility of the construction of embankments along the Kosi River because of sediment load, and recommended not to interfere with the river's natural flow through any structural means. The British government accepted the recommendations of the report at that time.

But after India's independence, the Bihar flood of 1953 was highly politicized, and Indian press media portrayed the Kosi river as 'sorrow of Bihar' and a 'serpent around the neck', as well as criticizing the government of India for neglecting the people of Bihar. Due to the public pressure, the government of India prioritized the flood protection programs in Bihar, and as a result, the Kosi Project was put forward.

The major activities in the Kosi Project were—to channelize and divert the Kosi River into a barrage, to control the Kosi River discharge into Indian territory, and to construct canals for irrigating the agricultural lands in India and Nepal. The Kosi Project was designed solely by the engineers and scientists at that time, and neither community consultation nor environment impact assessments were carried out. The project was initiated and based on a narrative that was external to the people living there; that recurrent floods from the Kosi river are damaging crops and impacting the lives of poor people living there, therefore the Kosi Project will mitigate the flood impacts and contribute towards prosperity in the region through better agricultural productivity. After the Kosi Project's completion in 1962, both Nepal and India intensified development activities in the Kosi flood plain, and, due to the false sense of security from the embankments, rapid urbanization and migration took place in the region (Mishra, 2008).

Flooding — helping hand or hindrance?

When I went to the field and talked to the older people who had settled there before the Kosi Project, they told me a completely different story. They said that floods were beneficial for them because floods carried nutrient-rich soil, which helped them with greater agricultural yield. They also mentioned that they had adapted their lifestyle to live with the floods, and they were not much bothered about them. They said floods would come but would quickly drain away in two to three days. They added that despite some inconvenience, the benefits of floods outweighed its impacts many times.

People mentioned that prior to the flood mitigation interventions, life was easier because they could grow much more rice and fruit in that area with a little effort in the field, and the foods they grew would last them for a whole year. In the past, there were excellent rice yields, lush fruit trees, and varieties of large fishes in the river. People were happy to live there, despite the flooding events. If any hazard exceeds the ability of the affected people or society to cope with their resources, it could be called a disaster, but it was not the case in the Kosi River basin. People knew precisely when the floods would occur and were also well prepared for it. Therefore, people living there did not regard floods in the Kosi River as a disaster.

Has the river been tamed?

After the completion of the Kosi Project in 1962, the embankment has breached eight times in the past (Figure 2) in the following years and locations: 1963 in Dalwa, 1964 in Jamalpur, 1971 in Bhataniya, 1980 in Bahurawa, 1984 in Hempur, 1987 in Samini, 1991 in Joganiya, and 2008 in Kusaha (Mishra, 2015). The Kosi River embankment breach in Kusaha was the costliest flood disaster in the History of Nepal and India (Dixit, 2009).

On 18 August 2008, the flood control embankments along the Kosi River in Nepal were breached due to the monsoon rain, and the river took a new course, shifting 120 km eastwards and flowing over settlements and inundating 3700 km² of land (Figure 3). The flood washed away 4648 ha of agricultural land in Nepal (UN-OCHA, 2009), and the river deposited seven feet of sand in the settlements and agricultural fields (Shrestha et al., 2010). Approximately 50 000 people in Nepal and 3.5 million in India were affected by the flood (Dixit, 2009) but the disaster of 2008 was not due to the excess flow in the river. At that time, the river flow was only about 166 000 m³s^-1, while the embankment can handle about 913 000 m³s^-1 of discharge. Several authors (Mishra, 2008; Dixit, 2009; Shrestha et al., 2010) have argued that river bed aggradation, use of improper technology, institutional dysfunction and governance deficits were the main reasons for the disaster.

The current situation

As suggested by Sadoff et al. (2013), if embankments confine the sediment-laden rivers like the Kosi, then sediment builds up raising the riverbed and causing embankment breaches (Figure 4). In the Kosi River, due to the riverbed's aggradation in some stretches of the embankments, the river flows at a higher level than the settlements, posing a threat to people living in that area. In response to that, local government in Nepal and India have raised and added layers of embankments to save infrastructure and settlements, which has further contributed to the drainage congestion and waterlogging in the area. Because of this, the locals claim that "scientists and engineers have turned the benign Kosi River into a disaster." People stated that flooding has become a severe hazard now and whenever the embankment breaches, the Kosi River dumps massive amounts of silt and sand on the paddy fields and makes it completely non-arable. Furthermore, because of the drainage congestion and waterlogging, there is an outbreak of water-borne diseases like cholera in every monsoon season. Also, the livelihood of farmers living there is at risk due to the declining agricultural productivity, which leads to insufficient crop produce to sustain their family throughout a year, forcing the farmers to work in other sectors, such as construction labourers.

What now? Trapped in one’s own net

Authors (Mishra, 2008; Dixit 2009; Sinha, 2009) argue that for the Kosi River, there is no increase in hazard intensity, but a significant increase in exposure and vulnerability, which is contributing to the overall flood risk in the region. Even while some (Dixit, 2003; Mishra 2008, 2015) argue that embankments are causing the problems in the Kosi floodplains, by blocking the natural flow of the river, there is no other easy alternative to replace or remove the embankments. That is because the local communities use embankments also as an emergency refuge during disasters as no other elevated areas are available in the floodplains (Figure 5). Therefore, embankments have created a deadlock situation in the Kosi river basin, where, on the one hand, embankments are causing problems and, on the other hand, functioning like lifelines during disasters. Hence the local governments regard investing in embankments as the only easy option available for flood protection and annually allocate budgets to strengthen and add more embankments.

 

Lessons learned

It is just a matter of time but, sooner or later, there will be another disaster in the Kosi River basin due to the embankment breach because the natural system of drainage has been altered there, and settlements were developed encroaching the space of the river. Reflecting on the debates around floods and developments, I propose that to be resilient against floods, firstly people need to change their mindset and learn to live with nature. Secondly, before designing any interventions, a thorough understanding of physical context (such as river dynamics and river characteristics), as well as social context (such as local people's livelihood and their interaction with nature, culture, and perception), is crucial. Otherwise, those intervention designed to reduce risk could have adverse repercussions like in the Kosi River, and the activities designed for 'common good' would turn into activities for 'common bad'.

References

Dixit, A. 2003. Floods and vulnerability: need to rethink flood management. In Flood Problem and Management in South Asia, 155–179. (Springer: Dordrecht.)

Dixit, A. 2009. Kosi embankment breach in Nepal: need for a paradigm shift in responding to floods. Economic and Political Weekly, 70–78.

Mishra, D K. 2008. Trapped between the devil and deep waters. South Asia Network on Dams Rivers and People (SANDRAP). (New Delhi.)

Mishra, D K. 2015. Sustainable engineering on disaster. Down to Earth.

Sadoff, C, Harshadeep, NR, Blackmore, D, Wu, X, O'Donnell, A, Jeuland, M, Lee, S, and Whittington, D. 2013. Ten fundamental questions for water resources development in the Ganges: myths and realities. Water Policy, Vol. 15(S1), 147–164.

Shrestha, R K, Ahlers, R, Bakker, M, and Gupta, J. 2010. Institutional dysfunction and challenges in flood control: a case study of the Kosi flood 2008. Economic and Political Weekly, 45–53.

Sinha, R. 2009. The great avulsion of Kosi on 18 August 2008. Current Science, 429–433.

Sinha, R, Gupta, A, Mishra, K, Tripathi, S, Nepal, S, Wahid, S M, and Swarnkar, S. 2019. Basin-scale hydrology and sediment dynamics of the Kosi river in the Himalayan foreland. Journal of Hydrology, Vol. 570, 156–166.

UN-OCHA. 2009. Kosi Flood disaster 2008. (United Nations Office for the Coordination of Humanitarian Affairs: Nepal.)

Vatra, K. 2013. Resilience-based approach to flood risk management in South Asia.

The ForPAc project has looked at how to usefully anticipate and preempt droughts and floods events. Here we look at their work on the recent extreme rainfall and flooding in 2018. Focusing in on Kenya, they dissected the nature, causes, impacts, and predictability of the floods in order to learn how we can improve flood risk management.

Similar to this year, the 2018 long rain season (March - May) was one of the wettest on record. Flooding is fairly common in Kenya; there is a major flood about every two years. The flooding was due to a series of multi-day heavy rainfall periods, rather than any one extreme rainfall day.

Meteorologically, they found that these multi-day rainfall events in 2018 resulted from atmospheric circulations associated with a combination of the Madden-Julian Oscillation (MJO), and the effects of distant tropical cyclones in the Indian Ocean. The MJO is a natural, periodic movement of rainfall around the Equatorial zone that effects East Africa (as well as many other places), and it so happened to occur at the same time there was tropical cyclone activity in the Indian Ocean. This provided the right conditions for moisture supply from the Congo basin to support big thunderstorms formation over Kenya and the surrounding area.

If we know the causes of the heavy rainfall, could they be predicted? The ForPAc project found that, yes, the MJO and tropical cyclone activity can be predicted on the subseasonal and short-term timescales. 'Great,' you might be thinking. 'We already know that we can predict rainfall, so what?'

It’s true, but we spend a lot of time focusing on the wrong type of forecast. The study authors indicated that seasonal forecast is typically not very good at predicting heavy rainfall that could lead to flooding during the East Africa Long Rains (unlike the more predictable Short Rains in Oct – Dec). However, the short-term and sub-seasonal forecasts actually can predict extreme rainfall like the kind we saw in 2018. In fact, ForPAc research shows that Kenya and East Africa is a ‘sweet spot’ for predicting at these sub-seasonal timescales. ForPAc is working with the Kenya Met Department to improve forecasts of rainfall and flood risk at these sub-seasonal timescales, which provide potentially quite useful lead times to prepare for flooding.

For example, during the 2018 long rains season, a number of people were marooned in flooded areas. The long lead sub-seasonal forecasts could have been used by the Kenya Red Cross Society to anticipate places that would be inaccessible due to flooding from the continuous rains and therefore support people to evacuate before the flooding.

Although it will take a lot more than just forecasts for extreme rainfall to mitigate these colliding risks, this is an important piece of that puzzle.

Reference

Kilavi, M, MacLeod, D, Ambani, M, Robbins, J, Dankers, R, Graham, R, Titley, H, Salih, A A M, and Todd, M C. 2018. Extreme rainfall and flooding over central Kenya including Nairobi City during the long-rains season 2018: causes, predictability, and potential for early warning and actions. Atmosphere, Vol. 9, 472.

It was also important for me to attend such meetings, not only to add value to my PhD, I am conducting research on impact-based forecasting of flood risks, but to get first-hand experience on the day to day lives of these communities and consequently how their livelihoods can be improved through research projects such as NIMFRU. The FAMVAC meetings have been taking place every month since September 2019 in the district to ensure that farmers in the district are well aware of the weather information, and to consequently help shape the planning of their farm activities to avert the impacts of floods on their livelihoods. In addition, experts in various matters including agricultural management, maternal health, and environmental conservation were involved in training the radio listener group members about various ways they can diversify their livelihoods. Topics covered include improved livestock and crops, establishment of tree nurseries, poultry keeping, sunflower production among others. The trainings would then be captured by a local radio station-ETOP radio to be broadcasted as per the timelines already established (Sunday at 2pm EAT and Monday at 10am EAT) to ensure that the information reaches the entire community.

The wise say that you learn more about people by interacting with them, hence I was very excited that I would be visiting the communities in Katakwi District. My journey started from Nairobi, Kenya on a fine Sunday morning. The journey from Kampala to Soroti started the following day, Monday 24th February 2020, with the whole team and we arrived at 2am Tuesday. The few hours of sleep and the fatigue did not however dampen the mood to start off the field trips to the villages of Anyangabella, Agule and Kaikamosing in Ongongoja, Magoro and Ngariam sub-counties respectively.

Having been in stakeholders’ meetings before I started my PhD, my initial thinking was that this will be just like a normal stakeholder set-up meeting, more of a facilitator-audience setup with presentations and questions. However, when we arrived at the first venue (Anyangabella village), I was surprised to find so many people; men, women, youth and even children. Secondly, the informal nature of the set-up and how sessions were designed was intriguing. There were approximately 82 farmers present from Anyangabella village. The sessions in the other villages followed the same set-up. I sat through the sessions listening and taking notes as I was keen to learn more about the villages and especially on their understanding of weather data. The facilitator Mr. Asalu from UNMA took the farmers through various sessions including development of the flood calendar for the previous season, interpretation of the next season weather predictions and a question forum session. I was also taken aback by how much farmers knew about weather information and how enthusiastic they were to learn more and more about what to expect in the coming season. I had an informal conversation with Ongongoja sub-county agricultural extension officer Mr. Omongole and he had this to note:

The FAMVACs has really impacted a lot on these people. At first, they were very ignorant of the weather information given. But after seeing that whatever is predicted is what happens, they now can make use of the weather information to plan their farm activities

For the next two days, we had the same set up village clinics at Magoro and Ngariam sub-counties with approximately 68 and 58 farmers present from Agule and Kaikamosing villages respectively. The same enthusiasm among farmers was also noted. For instance, while having a conversation with Rose, a farmer at Agule village, I asked her if the information has helped her in any way and she had this to say:

Floods have been devasting us every season. But since FAMVACs started, we do not fear the floods anymore because we now know what we need to do and what to plant. For example, in September the water was very much and hence we concentrated on digging furrows and planting rice. We are very grateful to FAMVACs and we wish that the project can be extended

The four-day field visits culminated with a stakeholder meeting at the Katakwi District office to present on the progress and finally in Anyangabella Village where the Farmer Voice Radio (FVR) Listener Group (LG) meeting was held. In Anyangabella, 12 LG members from the 3 villages were trained on other activities including establishing a tree nursery, sunflower production and how to handle the desert locust as ways to improve the livelihoods of the communities. These topics also formed the content for the ETOP radio to be disseminated to the whole community on designated time.

Upon reflection- I now understand why the radio program to disseminate the weather information and advisories to the community was named Acauta Akoriok which means ‘farmers get light’ - because for sure the farmers have improved knowledge on various aspects including weather and farming practices. Subsequently, seeing how the information can transform these farmers livelihoods, and for this case just weather information and advisories, I can imagine what more and more relevant information can do considering the existing information gap between the disaster responders and the affected communities. This has helped put my research on data and information for informed decisions on flood risks into perspective. The key message here is that research can only make a difference in the real world if we change from the traditional ways of doing research to a bottom-up approach, which means listening to the communities needs first before the design of research. And for sure NIMFRU project has been in the forefront to ensure such impacts are achieved in the communities. Kudos to the entire team!

SHEAR has been working to develop forecast-based early action in Kenya, Mozambique and Uganda, and Bangladesh. At this year’s Sharing Geosciences Online, we shared findings from a study to understand the different perceptions that stakeholders have about existing forecasting tools and data in Bangladesh, and about how these tools and data can be sued to support preparedness and response actions.

The study found that stakeholders need around 15 to 20 days of lead time to prepare effectively for floods, while current deterministic forecasts can only provide 2 to 3 days. This does short time frame does not enable stakeholders to assess and mobilize available resources or take action to protect communities from the impacts of flooding. In communities, while 2 to 3 days allows households to protect their homes and move livestock to safer ground, crucial livelihood decisions, such as when to transplant young crops into the field, are dependent on more time.

The study highlights the importance of the role of probabilistic forecasting, and of working with stakeholders to build understanding of and familiarity with probabilistic information to support flood preparedness decision-making for households, communities, and responding agencies.

Landslide early warning systems are a key measure to reduce these risks, and can provide authorities with the information they need to manage risk effectively and protect communities from the impacts of this hazard.

SHEAR is working in Nepal and India to develop and improve early warning systems for landslides, combining innovative technologies with local expertise to enhance understanding, management and response to landslide risk.

At Sharing Geosciences Online this year, we shared the preliminary results of a study which explores how to use the Internet of Things as a low-cost approach to providing early warning alerts in landslide-prone areas.

The BOULDER project has been working in the Bhote Koshi catchment northeast of Kathmandu, and area which is highly prone to landslides and was particularly affected by landsliding caused by the Gorkha earthquake in 2015.

Using trackers to tag boulders, boulder movement can be observed in real time. The trackers are triggered by movement, and automatically send updates to a central server. This data enables us to monitor how boulders move on different slopes, so we can better understand the risk they pose and keep of dangerous boulders.

Find out more about the BOULDER project's work, as well as other landslide-related research in Nepal and India that SHEAR is leading. You can also find out about how we are working to map and assess landslides.

Uncertainty

As a PhD researcher and teaching fellow at the University of Reading my life has been greatly impacted by COVID-19. My normal everyday lifestyle has been changed to one of abnormal chaos. What I am used to doing, saying, being and understanding has dramatically changed, almost overnight. I am now surrounded by uncertainty, anxiety and misinformation. For me, it is the uncertainty that I find most difficult to deal with. What will happen to my research? What happens if I fail to be productive? What happens if I don't pass my PhD? The list of questions goes on.

Mental Health

Mental health has always been an issue among the postgraduate community. Some of the conditions affecting students include isolation, imposter syndrome and depression [5] [6]. This is expected to be compounded during this pandemic. With the added stressors leading to anxiety, not just about the individual's health and their family's health but also the lost time and potential repercussions this may have on funding and submission dates and the potential loss of income [7].

I feel isolated – but I am not alone. All postgraduates and academics will feel this isolation to some degree. We are affectively being mandated to remain at home, to remain in isolation. Closed institutions, campuses, research centres and conferences mean that we are all now 'working from home' and going to conferences online [8]. What my 'working from home' looks like, will look very different for someone else.

Working from Home

Initially working from home seems very straight-forward. But there are so many challenges to face when doing so, and these challenges are varied and unique for everyone. Caring responsibilities, space issues and unsafe conditions are just a few challenges that many will come up against when transferring their work into the home [9]. And even if it happens, we are left questioning: what now? As researchers and academics our day-to-day work-life is diverse, made up of lots of different tasks: emails, skill learning, writing, reading, meetings, coding, data analysis, data collection etc. It is a never-ending list of processes that we fill our time with.

Now we need to find a new normal. I believe we now need to throw away the regular ways and methods of structuring our time and space, because for many of us home working is completely new. Reassessing what is considered normal, what is acceptable, what is realistic and more importantly what is manageable or possible is now of huge importance.

We can draw from the experiences of our peers, supervisors and other academics who have learnt to deal with chaos in their lives and still perform and highly achieve. There are many fine examples of those who are flourishing and successful at the University of Reading, and there will be others from partner institutions too. We all work differently and COVID19 will be affecting all of us. We all find that certain times of the day can be more productive. We all have different workspaces that we feel more comfortable and productive in [10] [11]. Let us find these spaces in our homes now.

To help do this we must start small and reflect continuously. Reflection is key to any process of growth. Using Gibbs' Reflective Cycle (Fig. 1) can provide us with a framework so that we go through the process in a structured way [12]. The University of Birmingham has a wonderful resource guide to starting reflective writing [13]. It poses key questions and methods that each of us can embrace on our own reflective journey. Each of our own institutions may have something different that will work for us. Taking this reflective time now may have a positive impact on our future working conditions and well-being.

Figure 1: Gibbs' reflective cycle. Starting from 'Description' to begin with, reflection can go through these different stages as either a single cycle or a multiple cycle [12].

There are many different people currently offering great ideas on how to find our new normal and how to try and change our planning and methods to adapt to the situation. From lifestyle coaches, celebrities, new applications for phones or even departmental emails – they are all leading us towards finding our own way to cope with, manage, adjust, and redesign our lives [14] [15] [16]. No one thing will work for everyone, because we are each unique. Now is the time to pick and choose what works for you. Collate some great ideas from numerous sources and chisel your own path forward.

For me, I think that we have this unique opportunity to examine the 24 hours of our day and gain insight into a 'natural norm' [17]. In doing this we can break down the things that we 'need' to do and the things that we 'want' to do. These things can be examined and reflected upon in depth. We can then action these, allocating the time appropriately. With practice we can reflect on our experience and make relevant adjustments. I believe that many of our highly performing peers have, through this trial and error methodology, come to an epiphany moment that the full 24-hour day is ours to master. Many of these peers have flexible working hours and I have received emails from them at what seems to be unusual times of the day. Once, I would have questioned their work-life balance, but now I think I feel a glimmer of understanding.

Help is at Hand

If this pandemic feels overwhelming, then where can we find support to help with this adaption and adjustment? It is ok to not be ok, and you can find a lot of help from many of the hotlines available in the UK and internationally to talk to someone impartial [18] [19]. The SHEAR Student Cohort have been putting in place peer support group sessions and check-ins for each other too, so that we still feel together, in the same situation, a kind of continuation of connection to others [20].

One of my favourite infographics is by Dr Zoe Ayres (@zjayres) as part of a #mentalheath series (Fig. 2). It takes the issue of not having a laboratory to conduct your research in and gives us many clear and excellent examples of things that we can do as PhD students to use our time away from the office in a constructive way [21].

Figure 2: Dr Zoe Ayres' infographic on actions to take if you are a PhD Student [21].

There is always so much pressure being placed on us as academics, we may find ourselves stuck in a cycle of guilt, anxiety and often depression. I do not want to add to this by saying that we can metamorphosise from this terrible tragic pandemic. However, we can focus on what is in our control and do our best in the current situation. Remember whoever you are, and wherever you are from, you are part of a community of academics and you are not alone. Reach out and talk to someone in your personal or professional circle, we are all distantly together. Be kind to others and to yourself.

References

[1] European Centre for Disease Prevention and Control (ECDC). (2020). Situation update worldwide, as of 7 April 2020. Retrieved from: https://www.ecdc.europa.eu/en/geographical-distribution-2019-ncov-cases [07/04/2020]

[2] World Economic Forum. (2020). Great Recession showed countries can't fight the coronavirus economic crisis alone. Retrieved from: https://www.weforum.org/agenda/2020/04/covid-19-coronavirus-economic-crisis-great-recession/

[3] Anderson, R. M., Heesterbeek, H., Klinkenberg, D., Hollingsworth, T. T. (2020). How will country-based measures influence the course of the COVID-19 epidemic? The Lancet. 395(10228). Pp 931-934. Retrieved from: https://doi.org/10.1016/S0140-6736(20)30567-5

[4] World Economic Forum. (2020). This is the human impact of COVID-19 – and how business can help. Retrieved from: https://www.weforum.org/agenda/2020/03/this-is-the-human-impact-of-covid-19-and-how-business-can-help/ [07/04/2020]

[5] Guthrie, S., Lichten, C. A., Belle, J. V., Ball, S., Knack, A., Hofman, J. (2018). Understanding mental health in the research environment: A rapid evidence assessment. Rand Health Quarterly. 7(3). Retrieved from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5873519/

[6] Sakulku, J. (2011). The Impostor Phenomenon. The Journal of Behavioral Science, 6(1). Pp. 75-97. Retrieved from: https://doi.org/10.14456/ijbs.2011.6

[7] UK Research and Innovation. (2020). Updated April 3: Impact of Coronavirus on WKRI-Supported research. Retrieved from: https://www.ukri.org/news/coronavirus-impact-on-ukri-supported-research/ [07/04/2020]

[8] European Geosciences Union: General Assembly. (2020). Vienna cancelled. Retrieved from: https://egu2020.eu/about_and_support/vienna_cancelled.html [07/04/2020]

[9] European Parliament. (2020). Press Release: COVID-19: Stopping the rise in domestic violence during lockdown. Retrieved from: https://www.europarl.europa.eu/news/en/press-room/20200406IPR76610/covid-19-stopping-the-rise-in-domestic-violence-during-lockdown [07/04/2020]

[10] You Matter. (2020). Homepage. Retrieved from: https://youmatter.suicidepreventionlifeline.org/ok-not-ok/ [07/04/2020]

[11] Samaritans. (2020). Homepage. Retrieved from: https://www.samaritans.org/ [07/04/2020]

[12] Gibbs, G. (1998). Learning by doing: a guide to teaching and learning methods. Further education unit. Oxford Polytechnic: Oxford.

[13] University of Birmingham. (2015). A Short guide to reflective writing. Retrieved from: https://intranet.birmingham.ac.uk/as/libraryservices/library/asc/documents/public/Short-Guide-Reflective-Writing.pdf [07/04/2020]

[14] Men's Health. (2020). Chris Hemsworth Shares a Killer At-Home Workout. Retrieved from: https://www.menshealth.com/fitness/a31977692/chris-hemsworth-home-workout-instagram/ [07/04/2020]

[15] The Metro. (2020). Where I work: Tony, a barefoot athlete and lifestyle coach who doesn't have any chairs in his home. Retrieved from: https://metro.co.uk/2020/03/20/work-tony-barefoot-athlete-lifestyle-coach-doesnt-chairs-home-12428598/ [07/04/2020]

[16] The Fabulous. (2020). Embark on a journey to reset your habits. Retrieved from: https://www.thefabulous.co/ [07/04/2020]

[17] Riddle, T. (2020). The Natural Lifestylist Instagram Post 03/04/2020. Retrieved from: https://www.instagram.com/p/B-fKOprnxcJ/?utm_source=ig_web_copy_link [07/04/2020]

[18] Workspace Design and Build Ltd. (2020). What can we learn from Google's offices about workplace design? Retrieved from: http://www.workspacedesign.co.uk/what-can-we-learn-from-googles-offices-about-workplace-design/ [07/04/2020]

[19] The Guardian. (2017). Anne Cassidy: Clocking off: The companies introducing nap time to the workplace. Retrieved from: https://www.theguardian.com/business-to-business/2017/dec/04/clocking-off-the-companies-introducing-nap-time-to-the-workplace [07/04/2020]

[20] Masefield, S. (2019). Use peer support to improve well-being and research outcomes. Nature. 572. Pp. 407-408

[21] Ayres, Z. (2020). Twitter Post: Scientist Without a Lab? 17/03/2020. Retrieved from: https://twitter.com/ZJAyres/status/1239983524259737606 [07/04/2020]

I attended COP to find out more about how disaster risk finance is being understood at the intersection of humanitarian response, disaster risk reduction and climate change adaptation. To do so, I followed the so–called 'loss and damage' (L&D) negotiations, formally known as the Warsaw International Mechanism, which addresses how climate–events including disasters and slow–onset events push people beyond the limits of adaptation to suffer both economic and non–economic losses. Disaster risk finance is increasingly being seen as part of the loss and damage conversation, as Action Aid's Harjeet Singh put it, "What we call loss and damage in climate parlance is nothing but humanitarian situations that are being created by climate change"... "In fact, this particular COP is largely for the humanitarian community." ¹

It was clear to everyone attending COP this year however that the backdrop to the conference was remarkable: waves of protests from the school strikes to Extinction Rebellion combined with evidence of a climate emergency that was clear from ongoing bushfires in Australia to record–breaking hurricanes in the Caribbean. Then, just weeks before COP was expected to begin in Santiago, social unrest spurred by inequalities led the Chilean President Sebastian Piñera to withdraw Chile's offer to host the conference. In their place Spain offered to host at the last minute in Madrid and all delegates, including myself, rapidly shifted their sights – and travel plans – to Madrid. It was a rocky start!

Despite this, for observers and delegates of the L&D negotiations, COP 25 was a particularly critical year. For one, it was the first year in which the 'WIM' mechanism was up for review. Secondly, after a year when climate impacts had hit home, perhaps more than ever before, there was hope that the widespread declarations of a climate emergency – an acknowledgement that mitigative measures taken up to this point are no longer enough to limit climate impacts – would be reflected through greater support for loss and damage compensation.

While L&D negotiations have been a feature of COP since 2014, until now coalitions of climate–vulnerable countries have been arguing for new and additional funding while the most developed countries have largely been able to withstand these calls, favouring an approach of risk assessment, evidence gathering and focussing on existing finance streams. The exception to this ongoing stalemate has often been climate risk insurance, which has been a rare site of consensus in the WIM, in part because of a support for insurance–based approaches among developed country governments, and in part because focussing on existing systems avoids the controversial topic of new and additional finance. Between negotiators from Japan, Norway and the US there was agreement that sovereign risk transfer facilities, such as CCRIF (the Caribbean Catastrophe Risk Insurance Facility), which mixes parametric insurance and regional risk pooling across governments, could be an opportunity for scale up. Meanwhile, the InsuResilience Global Partnership for Climate and Disaster Risk Finance and Insurance Solutions was also seen as a popular avenue for delivering finance that was particularly supported by Japan during the negotiations.

Insurance and risk transfer mechanisms can play a role in helping countries and individuals manage their risk to climate extremes, for example by identifying areas of climate risk that can signal areas of unsustainable development, for example where properties in risk areas become too expensive to insure (Jarzabkowski et al.2019). However, insurance also tends to be very expensive, and the pay–outs tend to be small as they are inherently constrained by the level of premium that countries can afford (Hillier, 2018). In the context of a climate emergency this may become a serious limitation as the hazards we are experiencing become ever more severe and frequent.

At COP this was limitation was highlighted during a bilateral meeting I joined with negotiators from the Bahamas, which was hit by Hurricane Dorian in early–September 2019. They were categorical about the crippling damage experienced in the wake of the hurricane, and while acknowledging that there are no silver bullets to cover this scale disaster losses, efforts to scale up financing were critical. Moreover, while the country received a pay–out from their sovereign risk insurance scheme (CCRIF) which provided liquidity in the immediate aftermath of the disaster, the total amount paid out was just $12.8 million USD. This pales into insignificance against the total cost of the disaster which was estimated to be $3.4 billion USD (Inter–American Development Bank, 2019). It was also striking that during the same week I was at COP, the UN Under–Secretary–General for Humanitarian Affairs and Emergency Relief Coordinator, Mark Lowcock, gave a major speech in London sounding a note of caution about insurance–based approaches, citing the experience of the CCRIF payout in the Bahamas in the wake of Hurricane Dorian as part of his concerns. ²

Considering this it would have been reassuring to see negotiators not just agreeing additional loss and damage funding (although some progress was made on the inclusion of paragraphs about the importance of finance in the agreed text ³), but exploring the detail of how systems could be made more robust to deliver this. For example, can sovereign risk insurance systems such as CCRIF be scaled up through new and additional funding, for example through donor contributions to premiums, to ensure their effectiveness in a world of more frequent and severe hurricanes? Or should we instead be looking to grant and aid–based approaches, or alternatives such as regional solidarity funds? Unfortunately what I heard at COP provided very little detail, nor did there seem to be any significant sharing of best practice between the humanitarian and climate sectors. Instead, political expediency won out over reflection about the effectiveness of the initiatives being proposed.

In summary, COP was a useful and productive week as a researcher. But as a citizen, COP25 failed to deliver action for those who need it most. For the most climate vulnerable countries, meeting the protection gap – the shortfall between the total cost of disasters and the finance currently received – will require not just new, and additional funding, but well thought through delivery mechanisms achieved through robust and not just expedient debate.

For more detail see full comment piece: Peter Newell & Olivia Taylor (2020): Fiddling while the planet burns? COP25 in perspective, Globalizations, DOI: 10.1080/14747731.2020.1726127

Hillier, D. (2018). Facing risk: Options and challenges in ensuring that climate/disaster risk finance and insurance deliver for poor people (Oxfam Briefing Paper). Retrieved from oxfamilibrary.openrepository.com

Inter-American Development Bank. (2019). Assessment of the effects and impacts of hurricane Dorian in the Bahamas. Washington, DC: IDB. Retrieved from idbdocs.iadb.org

Jarzabkowski, P., Chalkias, K., Clarke, D., Iyahen, E., Stadtmueller, D., & Zwick, A. (2019). Insurance for climate adaptation: Opportunities and limitations. Rotterdam and Washington, DC: Global Commission on Adaptation. Retrieved from www.gca.org

1. Why COP25 matters to the emergency aid sector
2. How data and innovative financing can help improve the world's response to humanitarian crises
3. Loss and Damage at COP25 – a hard fought step in the right direction

What data will we collect?

We will be collecting cores of sediments from a number of dams. This is part of our first work package, which aims to explore how cycles of droughts and flood propagate across the river basin and how they affect water resources distribution. Sediments provide key archives for such events and their impacts in the past.

Our coring work is novel — no one has collected or analysed these sediment records in the Limpopo before. These dams have been built in the last 50 to 80 years and not been 'disturbed' since then, either by dredging or by sampling. Thus, everything that happens in the sub-catchments is recorded in the sediments that accumulate at the bottom of the dams. So basically, studying those samples is like reading the story of the area for the past 50-80 years. This includes past floods, droughts and any event that might have indirectly affected ecosystems and communities that are relying on those dams for ready available water.

We will collect the samples from the bottom of the dams using a gravity corer. The chemical analyses of these sediments will shed light onto the sedimentation rate into each dam, the presence of layers deposed by floods, the sediments hydraulic properties and, eventually, on the long term records of floods and drought cycle, the effects of these in the distribution of pollutants, and the infiltration capacity towards the underlying aquifers.

In conclusion, it is worth the labour of building a raft to do this for the first time!

Getting ready to collect sediment cores

In order to collect the cores, we need to find a safe and stable way to go out into the dams. To do this, we built a special raft, more like a platform floating on plastic drums. The launch in the Bonkwakathako went smoothly so we are confident we will complete the surveys safely!

Taking it step by step to build our raft

Step 1 — preparation

Safety was our first priority, so that the raft would be as stable as possible. We started by drawing a 3D model of the raft based on the experience and drawings from our project collaborators from DABANE in Zimbabwe. The BSc students at BIUST did an awesome job with this! This was followed by a careful cutting and assembling process, which took a week, and involved some sore hands and arms!

Step 2 — finding some more equipment!

The next step was to find oars and life vests, which isn't an easy task in a landlocked country covered in deserts and bush! But here at BIUST we never give up...we found our equipment quite unexpectedly in a hunting shop!

Step 3 — building up our expectations

If we are lucky, we will be able to retrieve cores such as the one shown at the top of this blog (courtesy of previous work for the NERC-funded PULA project). This shows a flood event (or flood couplet) in the Notwane dam, in southern Botswana. Once you are able to pinpoint a flood event you can run some analyses and see if the flood has brought into the basin pollutants, for instance. And, if you compare the record of different dams in the same basin you will be able to see if extreme weather events such as cyclones and heavy rains are really affecting the whole basin or if there are local buffering factors instead. Measuring the permeability of these sediments also gives information on whether dam contributed to recharge the aquifer, through leakage.

To cut a long story short, sediments can tell you a lot about the environment. In other parts of our project, we'll be talking to village elders and residents about what these environmental changes mean for the communities who live around the dams, and how people cope.

The next step for us will be to characterise land use changes because this is a key factor that can enhance erosion and affect sediments yield and contaminates dispersal.

Step 4 — preparing our team

Before departure you also have to deal with the expectations of your students. Our Italian student, Florian, has this to say about the challenge of floating on a raft in Limpopo dams: 'For sure dealing with the wilderness of Africa will be surprising! I hope that hippos and crocodiles will be busy with their ordinary habits! I trust the work we have done so far, and I know the raft will carry us safely through the dams.'

Having tested the raft, the team is now in Zimbabwe, first stop Bulawayo and then Zhovhe dam (21°48'44.35" S – 29°42'40.83" E). We will proceed to the unnamed dam on the Bubi river (21°19'43.66" S – 29°56'9.85" E) (the wildest of them all!)