How earth observation data and analytics is helping tackle climate change
The environmental risks posed by climate change become clearer with each extreme weather event, but we are only starting to study the impact of these changes with an economic lens.
When researchers from the International Monetary Fund, the University of Cambridge, and the University of Southern California projected the long-term impact of temperature anomalies on global GDP per capita, they found that climate change could shrink the US economy by up to 10.5 percent by 2100. China is expected to take a 4.3 percent hit, while the European Union could witness a 4.6 percent reduction in real income over the next few decades.
Naturally, the debate over policy approaches to addressing climate change has become more vigorous among politicians and economists alike. And it is becoming evident that many of the core economic questions in the times to come are going to be climate-centric.
Consider some of the ways climate change could have a negative consequence on a country’s economy:
- Diminishing surface water and groundwater supplies from hotter weather decrease the crop yields in once-fertile areas
- Mounting losses from extreme weather-related events – right from increased wildfires in the US to flooding around Japan’s coats – threaten to bankrupt the insurance industry
- Extreme hot spells overload urban power grids, forcing electric utilities to spend millions of dollars in overhauling the energy infrastructure
- Hurricanes, made stronger by climate change, destroy roads and public property which must be rebuilt at great costs to the taxpayer
Improving resilience to climate change through space technology
Continually watching our planet from a distance, satellites generate a wealth of robust scientific data that empowers researchers to understand the impact of climate change. Satellites provide a birds-eye-view of what is happening on the Earth’s surface, both at a local, zoomed-in level and on a global scale.
The same data, when coupled with intelligent machine learning algorithms, can be used to unravel thematic and historical trends in Earth’s behavior across different areas and verticals, and especially climate change.
Earth observation is crucial to understand patterns of sea-level rise, surface temperature changes, rainfall variability, or even the cooling effects of increased stratospheric aerosols. For example, Sentinel-5P data can be used to measure methane, while water vapor measurements from MODIS allow us to understand weather patterns. These measurements are essential to create data-backed policies for climate action.
Knowledge transparency about the past and the present empowers us to prepare for the future as well. Earth observation resources form the basis for predictive modeling, allowing communities and businesses to plan sustainably, improving their resilience to climate change.
For example, farmers and governments can buffer the impact of drought by using real-time satellite data to monitor crop growth and development. Pan-sharpened imagery from SPOT/Pléiades can be used to calculate Normalized Difference Vegetation Index (NDVI), an invaluable resource to gain insights into crop health, yield estimation, and disaster assessment. When under-performing areas are identified before it’s too late, targeted application of fertilizers, pesticides, and irrigation can be taken up to mitigate the effect of incongruities in precipitation.
Similarly, insurers can take proactive actions to mitigate risk if they have actionable insights into their portfolio. Near-real-time weather data, when combined with environmental forecasting data from earth-observing satellites, allows insurance companies to map their exposed risk in a better manner. Further, they can use advanced analytics, such as the Automatic Image Anomaly Detection System (AIADS) algorithm, to detect temporal changes and gain the competitive advantage required to grow new ideas for offerings on risk indexes.
In the same vein, change detection algorithms can help utilities to make better decisions on infrastructure planning. Energy companies can leverage next-generation open earth observation data sources, such as Pléiades 1A/1B 0.5m resolution imagery, and Airbus change detection algorithms to measure changes at a specific location and thus prioritize their activities. Armed with better insights, utilities can optimize their outputs for predictive maintenance and forward-looking inventory planning.
Correspondingly, high-resolution satellite imagery and advanced analytics are invaluable tools for optimizing infrastructure management and monitoring critical projects, especially as cities rebuild after climate change-induced natural disasters.
Not only can satellite data help to identify the most urgent infrastructure gaps, but off-the-shelf algorithms such as Vasundharaa’s urban estimation or Pinkmatter’s settlement mapping can support more effective long-term infrastructure planning. The progress of the construction can also be monitored in a consistent and reliable manner through high-resolution imagery.
But not every city rebuilds. According to nonprofit group Oxfam, over 20 million people are forced to flee from their homes every year because of climate-fuelled disasters. To create disaster risk reduction plans and to manage response in times of need, governments and communities need evidence-based knowledge of the number of people displaced. Accurate, relevant, and timely satellite data can help with that.
Multispectral imagery from Sentinel-2 can assist with coastal area monitoring, inland water monitoring, glacier monitoring, and flood mapping. Deep learning algorithms from Aventior can detect apartments, houses, industrial buildings, and sheds in satellite images. And Orbital Insight’s truck detection algorithm can be used to track population movements for preparedness, evacuation, and return post-event.
Suffice to say, the range of environmental and socio-economic trends that can be examined using satellite imagery make space technology vitally important for climate change planning and mitigation. With some additional processing, spatial data can provide governments and businesses with the precise, cost-effective, and timely guidance they need to adapt to the effects of climate change.
How the Danish government used AI on satellite data to identify slurry tanks?
For the past few years, I heard a lot about the potential of Deep Learning in the geospatial industry to automate analytics and interpretation of raster data. There is a lot of research and some promising startups in that space, but there are not too many large scale success stories with production application of AI that come to my mind, so every such an example catches my attention.
An interesting project has been delivered at the end of the last year using Picterra platform. This Switzerland-based startup offers a cloud-based geospatial tool that enables users to train Deep Learning models on satellite and aerial imagery data without a single line of code. Picterra was used by the Danish agricultural advisory institute SEGES that has been tasked by the government to measure the level of the total country’s emission of ammonia.
The ammonia emissions happen mostly from so-called slurry tanks which are these large circular concrete structures where farmers gather all their animal waste. In addition, the emission is lower for containers having covered rooftops. As there were no data on the presence of the slurry tanks available and counting it manually on 34000 farms would take months, it seemed like a perfect use case for the object detection task.
SEGES had access to two critical data sources perfect for Deep Learning applications:
- A WMS imagery server covering the whole of Denmark at 25 cm of spatial resolution (1TB of data)
- Centroids of the 34.000 farms to be investigated
The first one was a great source of data with consistent quality over the entire country. The second one could be used to crop the whole dataset to areas of interest rather than to run the analytics over the entire dataset. The data has been plugged to Picterra, and the training datasets have been created.
Interestingly, Picterra builds their platform around the concept of low-shot learning which aims to deliver good results with a low number of labelled training data. The platform has a set of pre-trained models, training data augmentation workflows and a well-structured GPU architecture that allows you to apply transfer learning effectively and quickly build new classes of objects to be detected on the top of existing models. Once you have even a few objects of each class labelled, you don’t need to wait for hours to test the model, but you get the results within minutes to understand how much training data is still required.
FYI. If you have ever played with Deep Learning and you don’t have a fully automated data pipeline, you will understand how painful and time consuming the process of labelling, training and testing the models really is.
For the sake of this job, SEGES had prepared labels of just 56 slurry tanks with two classes of objects representing covered and uncovered reservoirs. Based on this input, the engine detected about 26k of slurry tanks with high levels of confidence (Recall > 90% and Precision > 85%). Based on the data the heatmap of emissions has be created.
The project is a great example showing that Deep Learning is already changing the way how geoscience is done. Until recently, the entry barrier to applying neural networks in our geoanalytics workflows has been too high. It required expensive GPU server setup, data scientists that would be able to develop data pipelines in a geospatial environment and tons of training data. With projects like Picterra, Deep Learning started to be accessible to the geospatial community… and guys from Picterra have just released a QGIS plugin to make it even easier for all of us.
Play around with the plugin and let me know your thoughts in the comments below.