Context <p>Decades of research have used spatial pattern metrics at time points, such as those in FRAGSTATS, to measure the net change of landscape configuration and composition during time intervals. However, metrics at time points fail to quantify a time interval’s gross change, which can be substantially larger than net change between two time points. The field of landscape ecology has lacked a conceptual framework and accompanying software to characterize patterns of gross change during time intervals in terms of patches that are spatially explicit with a typology that is mutually exclusive and collectively exhaustive.</p> Objectives <p>We present a new method and software named Spatially Explicit <b>Dynamic PA</b>tch <b>T</b>ransition <b>CH</b>aracterization (DynamicPATCH). DynamicPATCH reads a time series of raster maps of a Boolean variable to map, characterize, and quantify gross patch dynamics in ways that existing metrics miss. We design the method to compare time intervals that vary in their durations.</p> Methods <p>Our method identifies eight mutually exclusive and collectively exhaustive patch-based transition types: Disappearing, Appearing, Splitting, Merging, Perforating, Filling, Contracting, and Expanding. The method quantifies each type’s gross gain and gross loss in area. The method also shows how each type contributes to the gross increase and gross decrease in the number of patches during each time interval. We illustrate our method with a simple example as well as a case study on the dynamics of marshes and ponds in the Plum Island Ecosystems site of the National Science Foundation’s Long Term Ecological Research network (PIE-LTER) across the years 1938, 1971, and 2013.</p> Results <p>Our method reveals dramatic gross changes in the size and number of patches of pond and marsh at PIE-LTER, which traditional methods that report only net changes fail to capture. The transition types Perforating and Contracting account for 89% of gross loss in marsh size across the temporal extent. Annual gross increase in the number of marsh patches during the second time interval is 3.8 times greater than during the first time interval, with 40% of gross increase contributed by Appearing and 60% by Splitting. Gross gain of pond is more than twice the size of gross loss of pond, with Merging accounting for 47% of gross gain, followed by Appearing and Expanding. Appearing and Disappearing account for most gross changes in the number of ponds during the temporal extent, which is ten times greater than the net change in the number of ponds.</p> Conclusions <p>DynamicPATCH offers a new operationalized approach to characterize the spatial–temporal patterns of patch dynamics. DynamicPATCH can be applied to various landscapes to pursue important research agendas in landscape ecology, such as landscape fragmentation, habitat connectivity, process-pattern relationships, and more. Users can obtain the open-source Python package at <a href="https://github.com/zay1996/DynamicPATCH">https://github.com/zay1996/DynamicPATCH</a>.</p>