Tracking Low-Level Cloud Systems with Topology

TL;DR

This work introduces a topology-driven framework for tracking low-level cloud systems by leveraging merge-tree representations of cloud optical depth fields and partial optimal transport (pFGW) to match anchor points across time. Cloud objects are detected via COD superlevel sets, then grouped into cloud systems, whose trajectories are derived from probabilistic anchor-point matchings and system-level associations. Through marine and land case studies, the method demonstrates improved robustness to splitting/merging events and enhanced ability to capture evolving cloud structures compared with existing cloud-tracking tools and topology-based baselines. The approach enables detailed characterization of cloud lifecycles from satellite records and offers a scalable pathway for spatiotemporal cloud analysis with potential open-source release.

Abstract

Low-level clouds are ubiquitous in Earth's atmosphere, playing a crucial role in transporting heat, moisture, and momentum across the planet. Their evolution and interaction with other atmospheric components, such as aerosols, are essential to understanding the climate system and its sensitivity to anthropogenic influences. Advanced high-resolution geostationary satellites now resolve cloud systems with greater accuracy, establishing cloud tracking as a vital research area for studying their spatiotemporal dynamics. It enables disentangling advective and convective components driving cloud evolution. This, in turn, provides deeper insights into the structure and lifecycle of low-level cloud systems and the atmospheric processes they govern. In this paper, we propose a novel framework for tracking cloud systems using topology-driven techniques based on optimal transport. We first obtain a set of anchor points for the cloud systems based on the merge tree of the cloud optical depth field. We then apply topology-driven probabilistic feature tracking of these anchor points to guide the tracking of cloud systems. We demonstrate the utility of our framework by tracking clouds over the ocean and land to test for systematic differences in the two physically distinct settings. We further evaluate our framework through case studies and statistical analyses, comparing it against two leading cloud tracking tools and two topology-based general-purpose tracking tools. The results demonstrate that incorporating system-based tracking improves the ability to capture the evolution of low-level clouds. Our framework paves the way for detailed low-level cloud characterization studies using satellite data records.

Figures (27)

• Figure 1: Left: a 2D visualization of a scalar field $f$ with an embedded merge tree of $-f$. Middle: a 3D visualization of the graph of $f$. Right: an abstract visualization of the merge tree of $-f$. Local maxima are in red, saddles are in white, and the global minimum is in blue. The black contour passing through the saddle $a_3$ encloses two peak areas of the local maxima $a_6$ and $a_5$, respectively.
• Figure 2: Apply area-based simplification to stratocumulus clouds. From left to right: the COD field, the cloud area map, and the simplified map by excluding cloud objects smaller than 10 pixels.
• Figure 3: (a) A set of cloud objects enclosed by white contours; each contains a subtree of the global merge tree. Local maxima (a.k.a., anchor points) are in red, and saddles are in white. (b) Simplifying subtrees by removing the highlighted anchor points (inside green or yellow circles) and their parent saddles.
• Figure 4: Partial optimal transport. We match the merge trees of $-f_1$ and $-f_2$ (b) using using pFGW with $m=0.8$, producing a coupling matrix $C$ (c). Matched nodes in the two trees share the same color (a-b).
• Figure 5: Cloud system matching score. (a) Cloud system $X$ at time step $t$ and cloud systems $Y$ and $Z$ at time step $(t+1)$, along with their set of anchor points. (b) Selected rows and columns of the coupling matrix $C$.