Earth Observation (EO) and the applications of remotely
sensed imagery have rapidly begun to increase in volume in recent years. This is due to
the abundance of new large scale and smaller satellites (known as ‘cubesats’)
being launched, and a growth in the space sector due to the creation of private
businesses and access to large amounts of freely available data. Alongside this machine learning, online services and DataCubes are becoming more common. All of which is leading to huge developments in the democratisation of data - the way it is collected, processed and applied to create real societal change in support of the 2030 Agenda.
This abundance of data provides a colossal opportunity for
the data science community to investigate in order to help tackle the UN’s
Sustainable Development Goals (SDGs). The inaugural Data for Development Festival (2018), run by the Global Partnership for Sustainable Development Data, aimed to bring
together practitioners to whom data is a huge part of their work and show how
these data are being used to tackle the 17 goals and their associated 169 targets.
EO played a large role at the festival and it was
interesting to meet other people who are using similar data sources and
satellite imagery but looking at an entirely different goal, thus demonstrating
the versatility of the data. An interesting thought that I considered throughout the Festival after the 'Planet' stream plenary on Day 1, was that EO data can be viewed as the framework needed to coordinate further data collection in order to define, process and investigate specific indicators to support the SDGs successfully. I believe this to be important as EO data can only do so much on its own, we need other supporting networks in order to garner its full potential and tackle the goals.
There were examples of EO for the SDGs and their role in
tackling SDG 6 (Clean Water and Sanitation), 11 (Sustainable Cities and Communities) and 12 (Responsible Consumption and Production); as well as the indirect positive impacts
that the use of satellite imagery can have on the goals including 'No Poverty' (SDG 1), 'Zero Hunger' (SDG 2) and 'Good Health and Well-Being' (SDG 3). All of the research being conducted was fascinating
but what struck me most is that the majority of the work was very broad and
bidding to successfully tackle the entirety of an SDG and possibly impact on
others, whilst looking at small regions, predominantly on a national level. Applications of EO data for a number of sustainable development programmes were seen within Uganda, and a detailed analysis of development within Colombia by the DANE project was explored. The interconnected nature of the SDGs means that by tackling the issues
raised by one you are also helping solve others – thus unilaterally helping with
the UN’s targets. However, along with the general trend of EO much of the work
is still focused on the environmental impacts and supporting those goals which
are associated.
Other methods that are utilising satellite imagery and other
data sources have also begun to embrace machine learning techniques. I myself
presented in the ‘AI for Development: Case Studies from Tanzania and Beyond...’ session and demonstrated how open source
satellite imagery available via Google Earth has been used to help investigate
industries known to use modern slavery. Therefore, taking the opposite approach to
other research projects by focusing on a singular target rather than the whole
of the associated goal. Target 8.7
(part of Goal 8: Decent Work and
Economic Growth) aims to end all forms of child labour by 2025 and modern
slavery by 2030. This work is also looking at the ways in which machine learning can be applied to satellite imagery in order to locate brick kilns more efficiently. A further aim is to provide additional data to non-governmental organisations (NGOs) working on the ground to help eradicate contemporary forms of slavery, particularly debt bondage which is a common method of enslavement within the brick manufacturing industry which my work focuses on.
What I saw throughout the Festival were these two opposing approaches to using EO data for the SDGs, but methods that are ultimately quite complementary to one another. The approaches took both the form of broad brush strokes that tackle a whole SDG and can be applied to one nation, then in future work applied across different nations until there is a global assessment. Whereas others, including my own research, focus on a singular target but in a larger regional analysis. Other studies, such as population monitoring were using EO data at a global scale from the very start. It is important to understand the importance of scale for the goals and targets which are being assessed as only a very specific portion of a goal may be applicable to the use of remotely sensed imagery, but for the rest of the goal this may not be the case and other data collection methods and analysis may be more favourable. Although EO data may not be applicable for all cases it is important to remember that the more data that are made available on a global scale to tackle the SDGs, the more these data can be combined with smaller scale work assessing singular targets. A combination of methods is always better than one alone.
What I saw throughout the Festival were these two opposing approaches to using EO data for the SDGs, but methods that are ultimately quite complementary to one another. The approaches took both the form of broad brush strokes that tackle a whole SDG and can be applied to one nation, then in future work applied across different nations until there is a global assessment. Whereas others, including my own research, focus on a singular target but in a larger regional analysis. Other studies, such as population monitoring were using EO data at a global scale from the very start. It is important to understand the importance of scale for the goals and targets which are being assessed as only a very specific portion of a goal may be applicable to the use of remotely sensed imagery, but for the rest of the goal this may not be the case and other data collection methods and analysis may be more favourable. Although EO data may not be applicable for all cases it is important to remember that the more data that are made available on a global scale to tackle the SDGs, the more these data can be combined with smaller scale work assessing singular targets. A combination of methods is always better than one alone.
What was clearly demonstrated by the varied and pioneering projects that were presented in Bristol, is how interconnected our data collection methods should be in order to make the best use of EO data - including data collection from the ground, personal stories, census data and many more - as well as the interlocking nature of the SDGs themselves. Tackling one SDG provides an untold number of benefits for other goals. By working at different scales, with different data sources, on different goals and targets, Earth Observation is bound to provide an abundance of data that will be key to helping achieve the 2030 Agenda.
You can find out more about the Festival here.
Comments
Post a Comment