Peak Risk

This project will use open data, open science and open source to create risk models that can be used for environmental impact assessments and strategic planning.

There will be a focus on hurricane and cyber risk. However, the long term goal is to produce a model that can be used for comprehensive risk assessment and forecasting.

Based in Bermuda, the hurricane project will use sources of open data to help build a Bermudian hurricane model.

The techniques and software used will be translatable to other areas and should be applicable throughout the Caribbean.

The project also aims to develop open communities in Bermuda and to generate ties with others doing similar work elsewhere.

It is inspired by the magnificent work of the python community.

Why Peak Risk?

I grew up close to the Peak District, a national park in the UK, between Sheffield and Manchester. It was a great area for walking, hiking, climbing, caving, cycling and many other outdoor activities.

So we have peak. Naming projects after mountains and hills, that makes naming fun. For starters, there is Mount Snowdon. The highest mountain in Wales. But only one letter difference from Edward Snowden.

Cyber risk and privacy will both be a focus of the project.

The world is changing faster than at any time in my life. There has been a lot of change in that time. We are faced with global crises such as that of climate change. In many ways the world could be viewed as being at peak risk.

It is also a time of incredible opportunity for humanity. I believe that by free and open sharing of our knowledge, data and skills we can move the world in a good direction.

The python community has been a huge inspiration to start this project and thanks go out to all involved.

Models

Peak Risk will focus on local models, looking in detail at particular regions or risks.

However, the aim will be to work with open formats and standards and to help develop those standards.

Data

There are large amounts of very interesting data sets becoming widely available. A challenge is the sheer volume of separate sources of information and the rate at which it is growing.

There are some signs of consolidation and standards and software used across boundaries. Certainly, the same data science tools are being used in a wide variety of applications.

Cyber Risk

Snowdon is the highest mountain in Wales. It is also intended as a reference to Edward Snowden. There is a one letter garble, but these things happen.

It is not possible to understand cyber risk without an appreciation of the material that Snowden revealed to the world.

He has shown how completely broken cyber security is in the modern world. Laura Poitras' documentary, CitizenFour, winner of the best documentary Oscar in 2015, is an excellent account, filmed as Edward Snowden was in Hong Kong uncertain of his future. The film is also an excellent cyber security 101 guide.

There is currently a great deal of uncertainty about the future, which makes this a fascinating area for study. Cyber risk is a complex subject, how this world evolves will be determined by a complex mixture of legal, governmental, social and technical developments.

The whole area is also inextricably linked to privacy. There is a public debate about who should have access to just what data.

Increasingly, governments are introducing freedom of information acts. In the computer age, the costs of making information not only available, but also freely searchable are negligible.

Some governments are trying to use legal and other means to prevent the general public from having secure communications. So, the governments want to be able to collect our own communications but are less keen on transparency on their own actions.

It is clear there needs to be an informed, public debate about cyber security and privacy.

Within technology communities this debate is actively going on. The recent PyCon had several talks about how broken TLS (https) is, as well as efforts to fix the problem. There are many working on secure communication tools, such as Proton Mail. Others are working on distributed systems, in attempts to avoid central authorities which are always a ripe area for exploits.

Companies such as Google have responded admirably to the Snowden revelations, but are still under considerable government pressure to cooperate. Indeed, the Snowden docs showed how they were cooperating. Google were rightly angry when it was revealed that, notwithstanding this cooperation, the NSA was tapping into their internal networks to collect data as it was transferred from one site to another. Google responded by adding encryption to this communication.

Whatever you think of the actions of Edward Snowden, he has revealed to the world the truth of what is going on and allowed this debate to take place. It is in recognition of this that Snowdon will be the name for the Peak Risk Cyber Risk and Privacy project.

Privacy

As noted above, Privacy is inextricably linked to cyber security. Without strong cryptography individuals can have no real privacy.

Likewise, organisations which store personal data risk breaches of privacy unless they are able to properly secure that data. An organisation that does not have a healthy respect for individuals privacy is unlikely to do a good job with respect to information security.

The project will involve much data analysis and care will be taken to ensure that people's privacy is respected as that data is used. This can be a quite subtle problem when multiple data sources are involved. Individually, these data sources may not reveal personal information, but collectively they may do.

Hurricane

In October 2014 two hurricanes, Fay and Gonzalo hit Bermuda within a week. These hurricanes have generated much data that can be used to get a better understanding of how hurricane windfields interact with the topography and land cover. The aim is to use data such as satellite images and aerial photography to get a thorough understanding of the hazard.

Bermuda

The focus on Bermuda is simply down to that is where I currently live. The aim will be to develop local community and expertise and together work to build a better Bermuda for all that live here.

The small size of Bermuda makes it an ideal candidate for a detailed study of hurricane damage.

Other subjects of interest include:

  • why did two hurricanes 6 days apart follow such similar paths?
  • could the late intensification of Fay have been predicted?

Although a lot of data is available from these storms, finding it, cleaning it up and analysing it is a time consuming task.

Meanwhile, new data sources are constantly appearing. Weather Underground has a Personal Weather Station (PWS) project and has the following mission:

Our mission is to make the highest quality weather information
available to every person on this planet. Join our global community
and help us improve weather forecasting.

There are currently around 20 personal weather stations in Bermuda that are feeding data to Weather Underground. Some of these stations were recording data during the recent storms. Wind speed data is missing much of the time, presumably due to damage from high winds. Other stations appear to have stopped recording due to power loss.

Nonetheless, there is some useful data here. A related project is Open Weather Map. This project also has information on how to build your own weather station. This would be a great project for a school, with the added benefit of more data coming on stream.

Both these projects require signing up for an API key if you are making practical numbers of API calls to access data. It is hoped that historical data for research purposes can be obtained in bulk at little cost.

Environmental Research

Humanity is facing unprecedented challenges such as climate change, dwindling oil supplies and the reality that we cannot afford to extract much of the remaining oil without irreparable impact on the climate.

The project will create models that will allow environmentalists to assess the impacts of different climate strategies. It will also aim to develop mitigation techniques for areas that will be hit hard by the changes that we are unable to prevent.

Community Building

The project aims to build community around risk modelling. When humans work together, collaborate and openly share their knowledge and skills they are much more effective in solving problems.

There is a huge open science, knowledge and data movement in the world today, as exemplified by the python community.

The project will work to build up open science expertise here in Bermuda. It is a natural place for such research. It has a delicate ecosystem, surrounded by coral reefs. It is also uniquely placed in Mid-Atlantic as a location for gathering weather and climate data.

The ambitious goal is to build up an inclusive community of environmental expertise using free and open software, data and knowledge.

Development Philosophy

This project will build on, and work with, other open projects.

It will use open data sources and open science and open source.

The open world is moving very quickly. One of the biggest challenges for this project is to be able to complement work that is already going on. Just keeping track of all the new data sources and research is a challenging task.

Traditional, proprietary, risk models are certainly struggling to include new research and insights into their models in a timely manner.

Since the project is aiming to reduce death and destruction in the world there is a moral imperative to openly share data, research and knowledge.

Hence, this project will focus on working with others doing similar work and creating open data standards as well as open source software.

At the same time it will do local research, using advantage of local expertise and skills. Ideas will be generalised so they can be applied in other juristictions and we will actively seek to collaborate with others wherever they may be located.

Python

The python language and community has been an inspiration for many years. The recent PyCon 2015 in Montreal was the biggest ever PyCon with over 3000 attendees.

The community is actively working to be more inclusive and welcome new members. Many of the talks at PyCon were along this theme. All are available on Youtube and are recommended viewing for anyone wishing to get an appreciation of the power of collaboration in science.

Python is widely used in science, finance and education and will be the tool of choice for Peak Risk projects.

Naming Conventions

The organisation name, Peak Risk was chosen partly because it opens up the possibility of naming projects after hills and mountains.

Pull requests and other contributions

Contributions from others are more than welcome. This is a multi-disciplinary project. Indeed, one of the key challenges in creating comprehensive risk models is to bring together science and research from a wide range of disciplines.

Pull requests are always more than welcome.

Blog

The project has a blog