Modelling emissions from soil using EO data
Using TROPOMI and SMAP
Welcome to my blog, Earthbound! I started this newsletter as a graduate student last year. I was spending so much of my time reading papers and taking notes that I figured it would be great to share them with professionals outside of academia. This blog is perfect if you want to stay up to date with the latest earth observation research but don’t have the time to read through long, technical research papers. Subscribe to get emails about my newest posts.
Let’s talk about fertilizer.
Did you know that emissions from soil fertilizer use are a significant contributor to air pollution and global greenhouse gas emissions? I’ve spent a lot of my time focusing on traffic-related sources of air pollution, and I’ve recently started reading a lot more about greenhouse gas and air pollutant emissions from the agricultural sector. As the contribution of nitrogen dioxide from traffic and other man-made sources decreases, the proportion from croplands and the agricultural sector is expected to increase in the years to come. Measuring the emissions accurately will become increasingly important for meeting our net-zero goals.
Can we use remote sensing data to track soil emissions both spatially and temporally?
Well, a research team from the University of Michigan used TROPOMI nitrogen dioxide data to quantify NOx emissions from cropland in the Mississippi River Valley. They used something called a box model to quantify emissions from the soils in this area. They also mapped out “NOx pulses,” which is the sudden increase in NOx emissions after a rain event.
Link to paper: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GL089949
Research Goals for this study:
Estimate daily soil NOx emissions over the study area using two different data sources
Map out “pulses” in soil NOx emissions after a rain event
Data Sources:
1. NOx emissions: Tropospheric Ozone Monitoring Instrument (TROPOMI)
2. Soil moisture content: Soil Moisture Active Passive Satellite (SMAP)
3. Wind data: ERA4 reanalysis (a climate model)
4. Anthropogenic NOx: 2014 US National Emissions Inventory
High-Level Methods
Select an agricultural site that is not too close to an extensive number of anthropogenic NOx sources such as cars, airports, or nuclear sites.
Pick a reference site - this should be an upwind site that can be used to quantify background concentrations in NOx. This site changed for each daily measurement based on the wind direction that day. So if the winds were coming from the east, the upwind side would need to be on the eastern side of the study area to derive a background concentration.
Map nitrogen dioxide across the study area using TROPOMI and convert the data into nitrogen oxide concentrations using a ‘box model.’
Identify the change in soil moisture before and after precipitation events using the soil moisture volume product from the SMAP instrument.
Identify NOx pulses after a precipitation event by comparing NOx emission values with changes in soil moisture with SMAP
Conduct the same analysis using a Berkeley Dalhousie Soil NOx Parameterization (BDSNP) model to compare the results. This model is build off of observational and chamber-based measurements of soil NOx.
Major Takeaways
Satellite remote sensing data can be used to identify daily and seasonal fluctuations in soil NOx emissions
The satellite data was able to model the sudden increase in NOx emissions after a precipitation event. The peak NOx emissions occurred 1-3 days after the event.
The largest soil emissions are in the late spring and early summer, which aligns with the periods of high agricultural activity.
They conducted the same analysis over the same domain and time frame using the BDSNP model. What's interesting is that this model provides emissions values that are HALF of what we get from the satellite data, but the overall trends between the two are the same.
Remote sensing-based emissions are similar to chamber studies
Source
Huber, D. E., Steiner, A. L., & Kort, E. A. (2020). Daily Cropland Soil NOx Emissions Identified by TROPOMI and SMAP. Geophysical Research Letters, 47(22). https://doi.org/10.1029/2020GL089949
Thanks for reading! Next week, I’ll be writing an article on how COVID-related lockdowns may have increased the prevalence of wildfires in southern California.
If you enjoyed this post, feel free to share it:
Or subscribe: