Nepal Floods 2017: ‘We had the forecasts but not the system to link them with our preparedness and response’

Madhab Uprety, Disaster Risk Reduction consultant, and Sumit Dugar, Research Associate at Practical Action Consulting South Asia, reflect on efforts to improve disaster preparedness in Nepal.

Southern flood plains of Nepal have always experienced a varied degree of flooding during the monsoon.  Whilst humanitarian responders and district level agencies have always relied on a general level of preparedness based on average levels of risks leading to periodic preparedness interventions across these flood prone districts.

Rautahat-flood-1

Portion of the Lal Bakaiya dam that was swept away by a flood in Rautahat during August 2017. Source: The Himalayan Times.

So far most of the major river basins are equipped with early warning systems and there are existing protocols and mechanisms to communicate early warnings and plan for emergencies. Yet we seemed hardly prepared for the recent Terai floods that claimed hundreds of lives and rendered tens of thousands of people homeless.

The scale and extent of flooding was huge as incessant rainfall for two days and across the Churiya range unleashed floods across entire southern belt inundating 26 Terai districts.

Nepal-SDugar1

Picture credit: Discussion between community members during flood response exercise, by Sumit Dugar.

But the rainfall event was not off the radar of prediction and was indeed forecast several days in advance. The events were captured 10 and 3 days in advance by the global flood and weather forecast models. Though the flood alerts were issued by the Department of Hydrology and Meteorology (DHM) 24-hours in advance, disaster managers hardly move beyond their customary response approaches of rescue and relief.

Existing early warning systems in major river basins were instrumental in saving assets, lives and properties, but warnings and actions merely based on the real-time situation of rivers and rainfall were not adequate enough to cater every contingency, particularly in case of flash floods from small monsoon (intermittent) streams.

There is dire need of paradigm shift towards forecast-based emergency preparedness. Unfortunately, there is not any mechanism that links scientific forecasts within the existing humanitarian landscape.  And also it is largely unclear among the humanitarian actors and government stakeholders about what levels of forecast probabilities and magnitude are worthy to react.

Globally there has been a recent increased shift towards anticipatory preparedness and early actions based upon weather forecasts under the remit of Forecast-based Financing (FbF), a niche concept which envisions humanitarian agencies and government stakeholders making use of credible scientific information to anticipate possible impacts and mobilise resources automatically before a hazard event.

Nepal-SDugar2

Picture credit: Community member using hand siren during flood response exercise, by Sumit Dugar.

Practical Action together with support from the World Food Program have been advocating this approach in Nepal and the idea has already been put-forth across national Disaster Risk Reduction (DRR) platforms for discussion. FbF is currently being piloted in six flood prone districts of West Nepal and there is an intention to upscale these initiatives across entire southern flood plains of Nepal. Forecast thresholds and triggers have been identified and Standard Operating Procedures (SoPs) that includes science informed anticipatory actions have been developed and integrated in the local disaster preparedness plans.

However, regular communication of the forecasts and their effective interpretation among the end users constrains FbF implementation. The aftermath of recent floods in August 2017 have certainly increased relevance of FbF in disaster preparedness and response, and the pressing question now is whether the recent floods will act as a wake-up call and we start making forecasts part of our readiness or simply plan response and recovery in a normative business as usual manner.

The 2007 Floods: ten years of thinking big and acting early

Russell Burton is an expert in flood risk and incident management, Managing Director of RAB Consultants and a former flood warning team leader in the Environment Agency.  Russell looks back at changes in flood resilience since the 2007 floods.

I read John Curtin’s blog Flood risk management 10 years on – a journey of high and low tech improvements with great interest this week and found the list of improvements and the progress made impressive, especially against a backdrop of economic uncertainty and austerity.

abbey

Tewkesbury is situated in Gloucestershire and was one of the worst affected parts of the county when the floods hit in July 2007. Source: Panoramio

It also got me thinking back to my own experiences and how the changes are not only in the tech we use but also in the way we think and respond to floods. Like John, I have my own story of what I was doing on the day of the 2007 storm. In July 2007, RAB was one year old, so it was a time of optimism and hard work for me to get the company off the ground. On that day in July I had a meeting with Environment Agency Wales (remember them!) in Cardiff, to start planning a flood exercise of all things. I planned to drive from Lichfield to Cardiff via the M50 with RAB’s first ever employee as a passenger. To be fair the Met office issued severe weather warnings for the west midlands up to 2 or 3 days ahead of the storm but it would only be on the homeward journey for us and who was to say it would be that bad and in that location anyway? I blush at the thought of my gung-ho attitude but I think I reflect the mindset of the times, “Think maybe, act if we have to”.

So off we set and had a great meeting, but by the time we were heading home reports were already coming in of flooding and travel disruption. Nevertheless we made pretty good progress on the M50 till we were approaching the junction with the M5 at Stroud. By now the downpour was a monsoon, the motorway came to a standstill and gradually the road became a river as torrents of water flowed off the fields. The hard shoulder and lane one were impassable but after a two hour delay we were able to crawl past in the “fast lane”.

Motorway 2007

Strensham Services on the M5, July 20th 2007. Source: flickr

To be fair I wasn’t the only one “thinking maybe and acting if we have to”, at some point I probably passed the lorry full of flood barriers bound for Upton upon Severn but stuck on the other carriageway of the flooded M5.

So what’s changed? Personally, I take Severe Weather Warnings very seriously now and adjust my travel arrangements to suit. But 2007 was also a watershed in mindset for flood risk managers in the UK. For the Environment Agency, it was the birth of the mantra “Think Big, Act Early”. This simple phrase has stood up to challenge (Think big, Act Appropriately was one watered down suggestion I recall) and I believe it’s helped drive the high and low tech improvements listed by John. It’s made us ask questions like, how big, how shall we act and when is early enough? The Flood Forecasting Centre  quantify the scale and timing of the impacts through the Flood Guidance Statement, which now gives flood managers up to 5 days notice to decide how and when to act early. The new extended flood outlook offers a 30 day window for flood risk managers to prepare and plan the response. The 2013-14 floods (amongst others!) highlighted that Thinking Big didn’t always result in Act Early. There was a need to plan what early actions could be done with this extra lead time for unprotected communities, and so came Temporary Defence Deployment Plans and Major Incident Plans to set out the strategic and tactical actions flood managers have at their disposal. Although these are just a couple of examples, the other improvements in John’s list are also driven by the need to “Think Big, Act Early”.

BBC map

July 2007 floods at-a-glance. Source: BBC

So in summary, the high and low tech improvements that have been made over the last 10 years are all fantastic but for me the biggest change is in our mindset, to “Think Big, Act Early”. To bring this into sharp contrast, this week the Met Office predicted that Britain is heading for “unprecedented” winter rainfall after their new super computer predicted records will be broken by up to 30 per cent. Time to “Think Big, Act early”.

Forecasting and hazard early warning systems: international perspectives

Canada recently witnessed severe flooding with some river flows reporting their highest on record.  Whether this is becoming the norm or not following recent flood-rich years, delivering effective multi-hazard early warning systems is vital for enabling communities being prepared for future events.  The Sendai Framework for Disaster Risk Reduction, which was endorsed by the United Nations, encourages government agencies to support improved and effective early warning systems. So, how are agencies responding to such challenges of early warning for increased natural hazards?

Canadian-floods_2017

A man fixes a hose connected to a pump at his flooded house on Ile Bizard, Que., near Montreal, Monday, May 8, 2017. Source: Paul Chiasson/The Canadian Press

The European Geosciences Union General Assembly which took place between 23rd and 28th April in Vienna, was an opportunity to share progress and innovations in the field of forecasting and warning systems.  One PICO session brought together presentations on real-world case studies of system implementations, configured at local, regional, and national scales.

On emerging capabilities, Mikuličková presented on approaches to trans-boundary flood forecasting in Slovakia, where European scale predictions (EFAS) run alongside a local model.  Likewise, in Nepal, global predictions (GloFAS) are used in conjunction with regional Data-Based Mechanistic Modelling (DBM) techniques for community alerting.

Flood Outlook

The new flood outlook product for England and Wales. Source: Flood Forecasting Centre, Exeter.

Various approaches to flood outlooks were presented.  Pilling presented new methods for providing flood risk guidance for the month ahead in England and Wales which have been developed in response to the Winter 2015/16 floods. Thiemig presented the latest developments with the European Flood Awareness System (EFAS) which include seasonal outlooks, event hazard mapping and rapid risk assessments. 

For those that have no experience of running an operational flood forecasting system, the opportunity was provided to test yourself through the interactive game for pathways to designing and running an operational flood forecasting system.  This follows a series of interactive games developed by the HEPEX community over recent years at the EGU General Assembly.

PICO session

EGU PICO session on operational forecasting and warning systems for natural hazards: challenges and innovation.

Away from the PICO session, several other talks provided some insights into global approaches in flood forecasting.  Demargne presented work in France on hydrologic ensembles based on convection-permitting precipitation nowcasts for flash flood warning.  The new system, called Vigicrues Flash, follows major flooding which occurred in France in 2010 resulting in 24 casualties and 1350 people being evacuated.  Parry presented work on data availability, limitations, and the impact of observational uncertainty for operational flood forecasting in Scotland. Finally, Dottori gave an interesting talk on satellites, tweets, forecasts and the future of flood disaster management where the use of social media is being trialled to verify flood forecasting where no data exists.

For more information on our EGU presentation on research with the University of Dundee and SEPA, please read the recent HEPEX article on ‘public perceptions of flood warnings in Scotland.

Improving Disaster Management in Ghana

Michael Clark, Assistant Resilience Consultant from RAB, reports on efforts being made to improve disaster management in the West African country of Ghana.

In June 2015, significant flooding in Ghana led to the loss of many lives in Accra.  Four days of rainfall across the region resulted in many parts of the town being inundated by flood waters.  The flooding followed 185mm of rain in a single day – more than the average rainfall for the month.

Reports on the scale of the disaster state that Ghana’s Fire Service reported that a fire started at a bus station and then spread to a nearby gas station, causing an explosion which killed many people taking refuge from the flooding.

amediaagency-com

Flooding in Accra, Ghana (Source: Africa Media Agency)

Since then, the National Disaster Management Organisation (NADMO) have been taking steps to reduce the risk of future flooding.  Actions include educating members of the public about responding to the potential threat of flooding and removal of illegal structures built within waterways which had a key role in the 2015 floods.

Afri-Gate is a project led by the Bournemouth University Disaster Management Centre (BUDMC) and supported by RAB Consultants that aims to build resilience to both natural and man-made disasters in West Africa.  The initial phase of the project is currently identifying where there are challenges for disaster management in Ghana and will direct the second phase of the project which seeks to address how to manage and build the resilience of communities to a broad range of hazards including flooding.

I’ve recently returned from a trip to Accra to understand the disaster management and resilience structure in Ghana. I witnessed first-hand the huge strides being made by NADMO in establishing a robust disaster management structure and the credible steps taken towards more focus on disaster risk reduction through the establishment of a flood early warning system.

405306_10151203183082193_1966379880_n

Michael Clark is a Resilience Consultant with RAB in Durham.  Michael has worked in international development and was previously based in the city of Kumasi for five years with Thrive Africa.

It is anticipated that the first phase of the Afri-Gate project will conclude with a series of workshops later in 2017 where I’ll return to Accra along with colleagues from the research team at BUDMC in support of improved flood and disaster risk management.