Abstract In response to increasing temperatures and precipitation in the Arctic, ice-rich permafrost landscapes are undergoing rapid changes. In permafrost lowland landscapes, polygonal ice wedges are especially vulnerable, and their melting induces widespread subsidence triggering the transition from low-centered (LCP) to high-centered polygons (HCP) by forming degrading troughs. This process has an important impact on surface hydrology, as the connectivity of such trough networks determines the rate of drainage of an entire landscape (Liljedahl et al., 2016). While scientists have observed this degradation trend throughout large domains in the polygonal patterned Arctic landscape over timescales of multiple decades, it is especially evident in disturbed areas such as fire scars (Jones et al., 2015). Here, wildfires removed the insulating organic soil layer. We can therefore observe the LCP-to-HCP transition within only several years. Until now, studies on quantifying trough connectivity have been limited to local field studies and sparse time series only. With high-resolution Earth observation data, a more comprehensive analysis is possible. However, when considering the vast and ever-growing volumes of data generated, highly automated and scalable methods are needed that allow scientists to extract information on the geomorphic state and on changes over time of ice-wedge trough networks. In this study, we combine very-high-resolution (VHR) aerial imagery and comprehensive databases of segmented polygons derived from VHR optical satellite imagery (Witharana et al., 2018) to investigate the changing polygonal ground landscapes and their environmental implications in fire scars in Northern and Western Alaska. Leveraging the automated and scalable nature of our recently introduced approach (Rettelbach et al., 2021), we represent the polygon networks as graphs (a concept from computer science to describe complex networks) and use graph metrics to describe the state of these (hydrological) trough networks. Due to a lack of historical data, we cannot investigate a dense time series of a single representative study area on the evolution of the network, but rather leverage the possibilities of a space-for-time substitution. Thus, we focus on data from multiple fire scars of different ages (up to 120 years between date of disturbance and date of acquisition). With our approach, we might infer past and future states of degradation from the currently prevailing spatial patterns showing how this type of disturbed landscape evolves over space and time. It further allows scientists to gain insights into the complex geomorphology, hydrology, and ecology of landscapes, thus helping to quantify how they interact with climate change.