About this Post

I’m Steve Jessup, a meteorology professor at Columbus State University. I have been developing a tool for sounding analysis over the last year that I call the Meteorological Process Mapper (MPM), and this work will be presented at the AMS Annual Meeting in Denver this coming January. I was recently invited to present a seminar at the University of North Carolina Asheville to discuss the philosophy behind MPM and my progress in developing the tool so far. In this newsletter, I share substantial excerpts from this presentation with some written commentary for additional context.

What is MPM?

The Motivation and Central Philosophy of MPM

There is an art to understanding the atmosphere. The most successful forecasters and researchers have built a repository of intuitive insight, which they draw upon to make connections and anticipate consequences when they see patterns in the data they work with regularly. But much of this reasoning is implicit and qualitative. Could one build a tool that breaks down these data quantitative data, allowing one to construct parameters that are tied to physical atmospheric processes? Could such a tool be capable of objectively assessing relationships among these parameters, identifying which processes are more active, which might be suppressed, and how these processes might evolve, given what we can measure from a given environment?

This is the goal of process mapping.

The except below discusses where the idea of MPM came from, how my own experiences inspired its development, and the motivation that drove how it is constructed. For this and all videos in this post, I provide carefully edited subtitles.

Building a Process Mapper

While MPM is one specific tool, process mapping is a much more general philosophy with a wide range of potential applications across meteorology. MPM is designed to assess convective environments associated with severe weather and flash flooding, but one could use the same approach to build a similar tool built to forecast winter precipitation type, or perhaps to support wildfire forecasting, or any of a range of other applications. This next excerpt addresses how would one generally go about building a process mapper, and some of the key challenges I encountered as I built MPM.

A Sample Process: Precipitation Efficiency

The easiest way to explain how I’ve constructed MPM is by working through an example of one of its processes. Precipitation efficiency is a straightforward concept, but translating sounding data into a quantitative metric of precipitation efficiency requires accounting for several physical pathways in different layers of the atmosphere, each of which is determined by unique physical parameters and subject to its own constraints. In this excerpt, I explain why precipitation efficiency is an excellent candidate process by which to test process mapping, and how it is represented in MPM.

The Architecture of MPM

Process mapping involves not only defining individual processes, but defining the physical relationships between those constituent processes in a broader architecture that provides additional meaning. The current version of MPM is composed of 61 processes of different types that serve important functions. Each of those processes contribute uniquely to the superstructure that connects those processes to one another. This excerpt explains how MPM is built.

CAPE and the Buoyancy Regime

MPM quantifies convective available potential energy (CAPE) as more than a single number. It assesses positively and negatively buoyant layers and evaluates the connectivity of positive buoyancy throughout the profile, and provides a lapse rate fallback if no CAPE exists, as described in the excerpt below.

What MPM Does Well and What it Does Not

Forecasters use different tools for different purposes. Bulk sounding parameters, such as CAPE, precipitable water, storm relative helicity, and shear, are quite effective in anticipating hazards and storm intensity. Models and conventional sounding analysis are fairly good at anticipating convective mode: will a given environment produce supercells, a mesoscale convective system, or ordinary thunderstorms? While MPM is currently not particularly good at this kind of discrimination, it can supplement these approaches by identifying potential forecast busts and examining edge cases, as explained in the excerpt below.

Some MPM Results

An MPM Example: Major Flash Floods

Heavy rainfall and flash flooding might be the most straightforward hazard with which to begin evaluating MPM, because the two of key parameters in environments that produce heavy rainfall – low-level water vapor and moist ascent – are fairly straightforward to estimate, as explained in the excerpt below.

MPM Process Breakdown: Warm Rain Efficiency

As we have seen, precipitation efficiency in MPM is determined by different physical pathways: warm rain efficiency, mixed phase efficiency, ice growth structure, and the wet bulb zero transition zone. This excerpt summarizes how one of those sub-processes, the warm rain efficiency is represented.

Quantifying Ascent and the Updraft Budget

Several processes can inhibit the rate of ascent in an updraft, and for an updraft to form, a parcel rising from the boundary layer must overcome convective inhibition near the surface. MPM incorporates an updraft budget to represent vertical ascent in a fairly comprehensive manner, as explained in the excerpt below.

MPM Results: Version 1.0

This excerpt summarizes some early results from the first version of MPM, which are discussed much more extensively in another post, which you can read here. These early results show promise that MPM represents processes in a way that is physically intuitive and capable of detecting emergent patterns in the data.

Network Map Visualizations

By representing each process as a node and connecting those nodes when pairs of processes produce a statistically significant correlation, one can construct a network map. This excerpt describes the network maps for major flash floods before and after the heavy rainfall sets in, detailing how environmental conditions evolve from pre-conditioning to generating high rainfall rates.

Major Flash Flood Results: Regional Relationships

Processes computed in MPM are viewed with 3D rotating plots, with differently colored symbols representing cases in different regions. Alternatively, one could stratify the events based on their magnitude or intensity, convective mode, time of day, or even based on another process values, such as floods occurring in moister and drier environments. Some regional differences across the 30 major flash flood events and 80 tornado events are discussed in the excerpt below.

Lessons Learned in Building MPM

I closed the seminar by sharing how MPM handles missing data, which is critical when working with real data. Sometimes a hygristor fails in the mid-troposphere. I also share a few of my mistakes that I hope others will avoid if they choose to engage with process mapping, and close with future directions for process mapping.

Conclusion

If you’ve made it this far, perhaps you have an interest in trying process mapping for yourself. Feel free to reach out to me (my email address is on my CSU faculty profile page), and I’ll be happy to share my experience in working with MPM and discuss your ideas. I’ll also be at the American Meteorological Society Annual Meeting in January 2027 all week and would be happy to sit down and talk.

Author’s Note: This summary and the images contained herein are freely available for educational use. Please cite “Jessup, S. (2026). Inside the Meteorological Process Mapper: The Scientific Reasoning Behind the Framework” in scholarly work.

This work is made possible by those who believe in it. Thank you for considering a gift to the Earth and Space Sciences Foundation at Columbus State University to help sustain this research, support the publication of this and related work, and support student activities.

Please consider supporting this work.