Although the troposphere is commonly considered the closest layer of the atmosphere to the Earth’s surface, the planetary boundary layer (PBL) — the bottom layer of the troposphere — is definitely the part that the majority significantly influences weather near the surface. Within the 2018 planetary science decadal survey, the PBL was raised as an essential scientific issue that has the potential to boost storm forecasting and improve climate projections.
“The PBL is where the surface interacts with the atmosphere, including exchanges of moisture and warmth that help result in severe weather and a changing climate,” says Adam Milstein, a technical staff member in Lincoln Laboratory’s Applied Space Systems Group. “The PBL can also be where humans live, and the turbulent movement of aerosols throughout the PBL is significant for air quality that influences human health.”
Although vital for studying weather and climate, essential features of the PBL, reminiscent of its height, are difficult to resolve with current technology. Up to now 4 years, Lincoln Laboratory staff have been studying the PBL, specializing in two different tasks: using machine learning to make 3D-scanned profiles of the atmosphere, and resolving the vertical structure of the atmosphere more clearly to be able to higher predict droughts.
This PBL-focused research effort builds on greater than a decade of related work on fast, operational neural network algorithms developed by Lincoln Laboratory for NASA missions. These missions include the Time-Resolved Observations of Precipitation structure and storm Intensity with a Constellation of Smallsats (TROPICS) mission in addition to Aqua, a satellite that collects data about Earth’s water cycle and observes variables reminiscent of ocean temperature, precipitation, and water vapor within the atmosphere. These algorithms retrieve temperature and humidity from the satellite instrument data and have been shown to significantly improve the accuracy and usable global coverage of the observations over previous approaches. For TROPICS, the algorithms help retrieve data which are used to characterize a storm’s rapidly evolving structures in near-real time, and for Aqua, it has helped increase forecasting models, drought monitoring, and fire prediction.
These operational algorithms for TROPICS and Aqua are based on classic “shallow” neural networks to maximise speed and ease, making a one-dimensional vertical profile for every spectral measurement collected by the instrument over each location. While this approach has improved observations of the atmosphere right down to the surface overall, including the PBL, laboratory staff determined that newer “deep” learning techniques that treat the atmosphere over a region of interest as a three-dimensional image are needed to enhance PBL details further.
“We hypothesized that deep learning and artificial intelligence (AI) techniques could improve on current approaches by incorporating a greater statistical representation of 3D temperature and humidity imagery of the atmosphere into the solutions,” Milstein says. “Nevertheless it took some time to work out find out how to create the perfect dataset — a mixture of real and simulated data; we would have liked to arrange to coach these techniques.”
The team collaborated with Joseph Santanello of the NASA Goddard Space Flight Center and William Blackwell, also of the Applied Space Systems Group, in a recent NASA-funded effort showing that these retrieval algorithms can improve PBL detail, including more accurate determination of the PBL height than the previous state-of-the-art.
While improved knowledge of the PBL is broadly useful for increasing understanding of climate and weather, one key application is prediction of droughts. In keeping with a Global Drought Snapshot report released last yr, droughts are a pressing planetary issue that the worldwide community needs to handle. Lack of humidity near the surface, specifically at the extent of the PBL, is the leading indicator of drought. While previous studies using remote-sensing techniques have examined the humidity of soil to find out drought risk, studying the atmosphere may help predict when droughts will occur.
In an effort funded by Lincoln Laboratory’s Climate Change Initiative, Milstein, together with laboratory staff member Michael Pieper, are working with scientists at NASA’s Jet Propulsion Laboratory (JPL) to make use of neural network techniques to enhance drought prediction over the continental United States. While the work builds off of existing operational work JPL has done incorporating (partly) the laboratory’s operational “shallow” neural network approach for Aqua, the team believes that this work and the PBL-focused deep learning research work could be combined to further improve the accuracy of drought prediction.
“Lincoln Laboratory has been working with NASA for greater than a decade on neural network algorithms for estimating temperature and humidity within the atmosphere from space-borne infrared and microwave instruments, including those on the Aqua spacecraft,” Milstein says. “Over that point, we’ve got learned so much about this problem by working with the science community, including learning about what scientific challenges remain. Our long experience working on one of these distant sensing with NASA scientists, in addition to our experience with using neural network techniques, gave us a singular perspective.”
In keeping with Milstein, the following step for this project is to match the deep learning results to datasets from the National Oceanic and Atmospheric Administration, NASA, and the Department of Energy collected directly within the PBL using radiosondes, a variety of instrument flown on a weather balloon. “These direct measurements could be considered a type of ‘ground truth’ to quantify the accuracy of the techniques we’ve got developed,” Milstein says.
This improved neural network approach holds promise to display drought prediction that may exceed the capabilities of existing indicators, Milstein says, and to be a tool that scientists can depend on for many years to return.
