By the help of local data and mathematical models, Tina Comes is building the tools to help the world’s relief agencies in their decision making process.
This article is more than one years old, and may contain outdated information.
From the computer screen in her office in Grimstad, Tina Comes is building a tool that might be able to contribute to make as many correct decisions as possible, as quickly as possible, when a disaster occurs. The tool is in no way fully developed, but it is possible to see clear signs of what it will become.
“Most disasters are far too complex for a person to have a full overview and make good decisions every time. That is why we work on creating as good and operative tools to support decision making as possible,” Comes explained.
“In essence, it does not seem that difficult. You put in the data, process it and make a decision. Then you process the data again to see what consequences the decisions will have, and then you make a new decision,” she continued.
“But in reality, the decision makers are faced with huge amounts of information that might be uncertain, insufficient, conflicting, just plain wrong, contradicts other information or changes rapidly – especially during the first phases of a disaster and just after when the relief work is starting up.”
“At the same time, the results are dependent on the data that comes in. If the information data is not good, the basis for making a decision will not be good – even if the model is ever so good. So I might not be that easy after all.”
Because of this, Comes’ work has two main focuses. On the one hand she is concerned with developing and adopting good mathematical models that can process the data in the best way possible – where also time and precision are important variables. On the other hand, it is important to get an indication of what the consequences of the decisions will be.
“This is about money, but it’s even more important in other areas: Is the decision good and efficient compared to the number of people we reach through this action? And are the actions that are being brought out, feeding stations and field hospitals that are put up, emergency roads that are built, good actions where they are placed and carried out – not only on short term, but on long term as well?” asked Comes.
“When we know as much as possible about the consequence of a decision, the decision will be the best possible,” she said.
No more than 49 years ago, the world’s first research centre for disaster research was established at the state university in Ohio, USA. The intention was to provide systematic knowledge on several areas about how a local community, organisations and individuals can prepare for, handle and adapt to disasters and major destructive incidents.
During the first years, the main focus was on flood and tornado disasters in the US, but this soon changed to include disasters everywhere.
“During the last ten years, however, there has been a paradigm shift in the field, especially connected to globalisation and digitalisation. Earlier, the research had a strong base in the social sciences, but today, the focus is more on how technology can help. The transition to an increasingly digitalised society is therefore both a possibility and a challenge – which should be used in a constructive way. In my case by contributing to operatively good decision making tools,” said Comes.
The overall focus of her work is to build a bridge between formal analytical models on the one hand, and understandable and applicable evaluations on the other. Specifically, the goal is to design approaches, systems and tools that can help decision makers collect, process, communicate, share and evaluate data from several different sources and translate these data into information that they need to make informed decisions on the spot.
“The more complicated the situations are and the decisions have to be, the more exciting it is, I think. My driving force is my wish to make a difference, to contribute towards making the relief work after disasters better.”
The fact that the research is quite new also means that there are no easily accessible data or information to be used in an operative decision making tool. This applies to both descriptive data about what is done after a disaster and what mathematical models are used to process them – as well as information about situations as it is presented on maps in progress reports, official statistics, or subjectively via social media – as well as what needs the decision makers have: What they need to know to make the decision.
“Most of my work consists in finding old data. There is a lot of, but often unstructured, information available on the internet and from the individual relief organisations. This is being used. But it is important to also collect data directly from the disaster area. This way we can get the best results.”
Even though Tina Comes is in the lead of her research field, it is clear that she is not there on her own. In addition to drawing on collaborations with other researchers at CIEM and UiA, there is a big international collaboration. Among others, this includes UNOCHA; the UN secretariat that coordinates relief organisations efforts in disaster areas, and BFast Team; the Belgian Ministry of Foreign Affairs’ international relief organisation.
“I work with a network of researchers from all over the world. Research in this field is global,” she said.
For Tina, however, most of the international contact goes through the collaborative network The Disaster Resilience Lab. This is an interdisciplinary research group that connects researchers from the Netherlands, USA, England, Belgium, and Norway.
The researchers usually work separately, but when the typhoon Haiyan hit the Philippines in November last year, the members of the group went to the Philippines to collect data directly from the scene. Key questions like: What are the decision makers in the relief organisations doing to handle the enormous amounts of data available? How do the different actors and organisations share information? How do they perform risk analysis? And how do they use the available technology (smart phones, internet, etc.) in their relief work?
“We all had different areas and focuses. Mine was on logistics and the distribution of relief aid, where I am especially interested in hearing from the people on location how information about needs were actually converted into actions – how information on needs was realised in distribution of aid, and how aid from the outside affected the local markets and infrastructure in the time following the disaster.”
The international research group has also been to the Middle East to collect data on how the relief work in Syria takes place, and have applied for funds to go to West Africa to gather data on the relief work after the Ebola epidemic.
“Even though there are similarities, each disaster is different and unique. The same goes for relief work, which must be adapted to the need. Therefore, we wish for as broad a database as possible to acquire more knowledge that can provide better help next time a disaster happens,” said Comes.
During the time after her return, the data has been structured and prepared, and used in different mathematical models to see if new knowledge from an actual disaster can contribute to improve the operative decisions connected to the relief aid in future disasters.
Scenarios are central to the work on preparing the data. Helped by, among others, an iterative approach, where events and consequences are described over and over by mathematical models, it is possible to discover patterns and create an example where it is most likely better to place for example distribution centres and distribution sites for relief aid.
“The core, one could say, is that scenarios help the decision makers think about the implications of a decision before it is carried out. What we do, is make dynamic descriptions of the disaster’s development, including destroyed infrastructure, the need of relief shipments, available resources and decisions that have been made,” she said.
Because the work has an operative goal, Comes is not only working with researchers from other universities or public research groups. Professional ideal actors also contribute. Among others, the Grimstad based researcher cooperates with Map Action and Nethope.
“Map Action is a humanitarian foundation based in England who makes maps that – based on information from a disaster area – shows where the damages are most significant and the need for help is biggest. The intention is to make easily understandable map representations to be used by the relief workers on the scene. The focus is on user friendliness, in the same way I’m concerned with working towards an operative decision making tool. So here we have some common ground.”
“Nethope is an umbrella organisation, based in the US, for voluntary relief organisations that contribute by having their members share knowledge about technological communication. In a disaster situation, there is often a network overload – the bandwidth is just not good enough for all the information being sent. Then it’s about evaluating what information is more important and make space for that. Here too, the organisation and I have some focus point in common and cooperate” she said.