<p>The freight system serves as the fundamental support for urban logistics, in which heavy trucks constitute the majority of the flow of goods. Hence, the flow of heavy trucks can be regarded as a proxy for analyzing dominant freight activities across urban zones. In this paper, we use heavy truck flow data to classify the freight activity patterns of four case cities in China. First, we partition each city into zones and construct a freight correlation network in which the zones serve as nodes and the correlations between their freight time series serve as weighted edges, to capture the similarity of freight activities across zones. Next, we apply a community detection algorithm to divide these zones into clusters with similar time series. We focus our analysis on the four largest clusters, which collectively encompass more than 97% of all zones across the studied cities. Among these clusters, there are significant differences in the temporal and spatial patterns of heavy truck freight. In accordance with the temporal activity pattern of the freight time series within each cluster, we identify these four clusters as morning, daytime, evening and nighttime patterns. Moreover, we show that freight activity patterns meaningfully conform with urban land use, which validates the use of freight data as an effective proxy for understanding urban functional structures. This approach bridges a gap in existing research by introducing a data-driven approach for identifying freight activity patterns and provides valuable insights and potential references for urban land use planning and freight policy formulation.</p>