How weather patterns impact child mortality
PLUS: predicting heavy metal pollution, identifying green infrastructure locations, and new indoor air quality datasets.
Hey guys, here’s this week’s edition of the Spatial Edge — a weekly round-up of geospatial news that you can digest in less than 5 minutes. It’s almost as efficient as a Cloud Optimised GeoTIFF.
In today’s newsletter:
El Niño and Child Mortality: Higher child mortality from El Niño temperature increases.
Predicting Soil Pollution: Using ML to predict soil heavy metal pollution hotspots.
Protecting Coasts with Nature: Best green infrastructure spots for flooding.
Indoor Air Quality in India: Measurements from 30 indoor sites over six months.
Maxar's WorldView Images: High-resolution images at 30cm resolution.
Research you should know about
1. How El Niño weather patterns affect child mortality rates
A new paper in Nature Communications finds that temporary increases in temperature (from El Niño Southern Oscillation events), can lead to higher rates of child mortality. The study was conducted across 38 low and middle-income countries.
This happens because El Niño can cause extreme weather events like heavy rains, floods, and droughts. This in turn can:
disrupt food supply, increase the spread of waterborne diseases, and strain health resources
lead to malnutrition, dehydration, and increased vulnerability to infections, due to the stress of dealing with these extreme weather conditions
reduces access to healthcare, clean water, and sufficient nutrition, further increasing the risks to children’s health
Pregnant women who are exposed to these conditions face higher risks of child mortality. The effects were larger in rural areas and among populations with lower education levels and unsafe drinking water sources.
The article used geospatial data from NOAA, and the Demographic and Health Surveys (DHS).
P.S. the code has been made open source: https://zenodo.org/records/11408167
2. A new method of predicting soil heavy metal pollution using geospatial data and deep learning
A study in Scientific Reports presents a new method for predicting soil heavy metal pollution using Graph Neural Networks (GNN). Soil heavy metal can massively impact human health, and soil quality (and therefore crop yield), and can disrupt ecological health.
Traditional methods of predicting soil heavy metal pollution don’t account for spatial autocorrelation. In other words, they treat each data point (like a soil sample) as if it were isolated from others. However, in reality, soil heavy metal pollution in one patch of land is impacted by soil heavy metal pollution in a neighbouring patch.
To address this, the authors propose using GNNs to predict pollution hotspots in the Pearl River Basin, an area in Southern China where pollution is a concern due to industrial and agricultural activities.
The model was trained on 142 surface soil samples taken from the Pearl River Basin. Their model outperformed other models (e.g. Support Vector Regression, Random Forest) with an R² of over 0.8.
3. Using nature to protect our coasts: finding the best spots for green infrastructure
A new paper funded by NASA develops a methodology to identify which areas are most suitable for implementing green infrastructure (GI). GI involves using ‘nature-based’ solutions to handle problems like stormwater runoff, flooding, and urban heat. For example, Rain Gardens are planted areas designed to absorb and filter rainwater.
The researchers combine remote sensing data and geospatial modelling. They use a bunch of different data like Sentinel-1, Sentinel-2, Landsat, and LiDAR, to come up with a Green Infrastructure Suitability Index.
They identify suitable areas for GI applications using criteria like slope, water table depth, and soil drainage class. They ultimately find that 91% of identified suitable areas for GI are residential, which means there’s a pretty high potential for integrating GI in urban settings.
Geospatial Datasets
1. Sentinel-1 burst archive from ASF
ASF’s Sentinel-1 burst archive is complete and will be continuously updated.
Sentinel-1 captures detailed radar images using two methods: Interferometric Wide (IW) and Extra-Wide Swath (EW). These methods use a scanning technique called Terrain Observation with Progressive Scans SAR (TOPSAR). In TOPSAR, the radar antenna scans an area and then switches to the next adjacent area. During each scan, it sends out quick bursts of radar pulses, which essentially create continuous snapshots of the ground.
These burst-based products are great for getting data in small areas and for projects that need precise alignment of data over time. Precise alignment basically means that each radar burst is taken in such a way that it matches up perfectly with snapshots taken at different times.
2. Lightning dataset from NASA
NASA has released a new lightning dataset. The Geostationary Lightning Mapper (GLM) uses a system called CIERRA (Cluster Integrity, Exception Resolution, and Reclustering Algorithm) to organise lightning data.
CIERRA breaks down the information into levels: individual lightning events, groups of nearby events, patterns formed by these groups, flashes made up of multiple events and groups, and the overall areas where these flashes happen. This helps us better analyze and understand lightning patterns.
3. Indoor air quality dataset in India
A new dataset on spatiotemporal measurements of air quality in India, covering 30 indoor sites over six months during summer and winter seasons. It contains various types of indoor movements from apartments, classrooms, laboratories, and more.
Other useful bits
Maxar has just released images from its new WorldView Legion satellites, which provide images at an insane 30cm resolution.
Not really geospatial news, but a hilarious new paper uses ‘LinkedIn cues’ to predict narcissism and intelligence. Cues for narcissism include not smiling in a profile pic, including public speaking as a skill, and mentioning leadership roles in the Experience section.
ESA has released the Space Environment Report for 2024. Some of the key takeaways:
Orbits are becoming overcrowded with debris from defunct satellites and rockets.
Although mitigation efforts are improving, they’re insufficient to stop the increase in space debris.
ESA wants to reduce debris through its Zero Debris Approach by 2030.
A record number of satellites were launched in 2023, which essentially increases congestion in low-Earth orbit.
ESA also added new features to its Heritage Data Visualisation (HEDAVI) tool, which helps us easily visualize Earth observation data.
They developed models to measure and predict thermal stress and other climate-related changes.
This involved creating an AI-based model to predict thermal stress down to individual streets, developing a method for optimizing tree placement to reduce heat, and investigating the impact of local winds.
A Python package (scikit-eo) for satellite remote sensing data analysis has been released, which provides access to the most commonly used Python functions in remote sensing analysis for environmental studies.
Jobs
Our World in Data is looking for a Research and Data Project Lead in Economics.
Canva is looking for a Technical Engineering Manager based in Australia/New Zealand.
Global Facility for Disaster Reduction and Recovery (GFDRR) is looking for a young Geospatial Data Scientist who will join the World Bank’s Junior Professional Associates (JPA) Program.
Coastal Carbon is looking to fill a number of vacancies: AI Scientists, AI Engineers, MLOps Engineers, and Chiefs of Staff.
UNDP is looking for a GIS Associate and GIS Developer in New Delhi.
SpaceKnow is looking for a Machine Learning Engineer for Geospatial AI.
GAF Munich is looking for a Project Manager in Earth Observation and Copernicus Land Cover and Land Use (LCLU) downstream Services.
Development Seed is recruiting a Geospatial Services Engineer to develop geospatial data services and advance open-source geospatial technologies.
The United States Geological Service is looking for a fellow specialising in remote sensing of agricultural practices.
That’s it for this week.
I’m always keen to hear from you, so please let me know if you have:
new geospatial datasets
newly published papers
geospatial job opportunities
and I’ll do my best to showcase them here.