Environmental communication strategies in green consumption: spatiotemporal shifts across six domains revealed by social big data

Abstract

The dissemination of green consumption information is essential for promoting sustainable behaviors and guiding environmental policymaking. However, the underlying dynamics of its dissemination across social networks and geographic regions remain insufficiently explored. In this study, the spatiotemporal evolution of green consumption information dissemination on social media was analyzed, focusing on six key aspects: clothing, food, housing, transportation, products, and tourism. A comprehensive analytical framework was constructed by integrating network and spatial perspectives. Social network analysis and exploratory spatial data analysis were applied to data collected from Sina Weibo, China’s leading social media platform, enabling an investigation into the multifaceted nature of green information dissemination. The results reveal that green consumption social networks exhibit clustering and tightening trends, while maintaining random network properties. Key nodes, particularly news media and environmental protection accounts, act as influential centers. Spatially, the network displays small-world characteristics, with Beijing, Shanghai, Guangdong, Jiangsu, and Zhejiang serving as critical information hubs. Additionally, geographical imbalances in information dissemination are evident, with high-value clusters concentrated in economically developed regions. Based on these findings, three targeted strategies are proposed: a network optimization-oriented guidance strategy to enhance dissemination efficiency and structural connectivity; a community integration-oriented guidance strategy to activate grassroots engagement and diversify participation; and a regional linkage-oriented guidance strategy to reduce spatial imbalances and promote cross-regional synergy. This research provides valuable insights for policymakers and stakeholders in promoting green consumption practices and environmental awareness across diverse areas.

Appendices

Appendix 1

See Table 5

Table 5 Network characteristics of green consumption social networks under different K-cores during the green policy preparation stage

See Table 6

Table 6 Network characteristics of green consumption social networks under different K-cores during the deepening stage of green transformation

See Table 7

Table 7 Network characteristics of green consumption social networks under different K-cores during the comprehensive advancement stage of green consumption

See Table 8

Table 8 Weighted degree of green clothing social network

See Table 9

Table 9 Betweenness centrality of green clothing social network

See Table 10

Table 10 Weighted degree of green food social network

See Table 11

Table 11 Betweenness centrality of green food social network

See Table 12

Table 12 Weighted degree of green housing social network

See Table 13

Table 13 Betweenness centrality of green housing social network

See Table 14

Table 14 Weighted degree of green transportation social network

See Table 15

Table 15 Betweenness centrality of green transportation social network

See Table 16

Table 16 Weighted degree of green products social network

See Table 17

Table 17 Betweenness centrality of green products social network

See Table 18

Table 18 Weighted degree of green tourism social network

See Table 19

Table 19 Betweenness centrality of green tourism social network

Appendix 2

See Table 20

Table 20 Weighted degree of green clothing social network

See Table 21

Table 21 Eigenvector centrality of green clothing social network

See Table 22

Table 22 Weighted degree of green food social network

See Table 23

Table 23 Eigenvector centrality of green food social network

See Table 24

Table 24 Weighted degree of green housing social network

See Table 25

Table 25 Eigenvector centrality of green housing social network

See Table 26

Table 26 Weighted degree of green transportation social network

See Table 27

Table 27 Eigenvector centrality of green transportation social network

See Table 28

Table 28 Weighted degree of green products social network

See Table 29

Table 29 Eigenvector centrality of green products social network

See Table 30

Table 30 Weighted degree of green tourism social network

See Table 31

Table 31 Eigenvector centrality of green tourism social network

Appendix 3

During the green policy preparation phase (Fig. 10), the green food social network exhibited a pronounced high-high clustering pattern, concentrated in Central China (Henan, Hubei) and East China (Jiangsu, Anhui, Zhejiang), reflecting continuous geographical distribution. Low-low clustering emerged in Gansu and Tibet, aligning with cold spots, indicating lower discourse volumes than neighbors. As the green transformation progressed, Shanxi and Tianjin emerged in low–high clustering zones, indicating less green food discussion than neighbors. However, Shanxi's position within a hotspot suggests a relatively high overall discussion volume in the region and its surroundings, possibly indicating a growing local interest and discourse on green food due to specific geographical factors. In the comprehensive GC advancement stage, high-high clustering expanded to encompass Central China and parts of East China (Jiangsu, Anhui, Shanghai, Fujian), mirroring hotspot areas. Meanwhile, low-low clustering persisted in Xinjiang, Qinghai, and Tibet, reinforcing their backwardness in green food discourse, in line with cold spot patterns.

Fig. 10

Spatial autocorrelation and hotspot analysis of green food social network

Figure 11 reveals that during the green policy preparation, high-high clustering areas of the green housing social network were mainly located in North China (centered on Beijing and Hebei), East China (represented by Shandong, Jiangsu, and Shanghai), and Central China (with Henan and Hubei as key areas). Shanxi and Anhui, though low–high clustered, benefited from high discourse in their vicinities, showcasing interest. Qinghai and Tibet, as low-low clusters, lagged behind in discussions, aligning with cold spots. As the green transformation progressed, Shanxi and Anhui transitioned to high-high clusters, reflecting a surge in green housing interest due to spatial diffusion and policy impetus. In the comprehensive GC stage, high-high clusters further expanded southwards, with Hunan and Jiangxi emerging as new hotspots, signifying a broadening geographical reach of green housing discourse.

Fig. 11

Spatial autocorrelation and hotspot analysis of green housing social network

Figure 12 illustrates that during the green policy preparation stage, green transportation hotspots clustered in North (Beijing, Hebei), East (Shandong, Jiangsu, Shanghai), and Central China (Henan, Hubei). As the green transformation deepened, high-high clusters dwindled, notably with Shanxi shifting to low–high, signifying lower discourse than neighbors. Tibet emerged as a low-low cluster, highlighting limited green transportation discussion. With GC policies'comprehensive promotion, high-high clusters expanded southwards, with Hunan, Jiangxi, and Zhejiang emerging as new hotspots. This shift underscores southern China's growing attention to green transportation, mirroring the region's advancing GC culture.

Fig. 12

Spatial autocorrelation and hotspot analysis of green transportation social network

Figure 13 depicts that during the policy preparation stage, green products social network hotspots centered in Central China and East China (Jiangsu, Shanghai, Zhejiang, Fujian). Anhui and Jiangxi, as low–high clusters, showed potential growth amidst high-discourse neighbors. Xinjiang, Tibet, Qinghai were low-low clusters, mirroring cold spots for green product discussions. As green transformation deepened, high-high clusters expanded north to Hebei, Shanxi, Tianjin, indicating northward spread. Tibet remained the sole low-low cluster, hinting at discussion growth elsewhere. In the GC advancement stage, hotspots shifted to Henan, Hubei, Anhui, Jiangsu, Shanghai, maintaining high interest. Shanxi slipped to low–high, reflecting stagnation, while Qinghai newly joined low-low ranks, signaling a decline in green product discourse.

Fig. 13

Spatial autocorrelation and hotspot analysis of green products social network

Figure 14 reveals that during the green policy preparation stage, high-high clustering areas of the green tourism social network are in East China, represented by Jiangsu, Shanghai, and Shandong, as well as in Central China, represented by Henan and Hubei. The low–high clustering area is Anhui, while the low-low clustering area is Tibet. The spatial clustering distribution was similar to hot and cold spot areas. In the deepening stage of green transformation, high-high clustering areas remained stable with no significant changes, indicating that previously formed hotspot areas continued to maintain their activity in disseminating the green tourism theme. Ningxia became a new low–high clustering area, suggesting that although Ningxia’s discussion volume on green tourism was not as high as neighboring high clustering areas, its discussion activity relatively increased at this stage. No low-low clustering areas appeared at this stage, possibly suggesting a general increase in attention to the green tourism theme. In the comprehensive advancement stage of GC, the range of high-high clustering areas contracted, mainly concentrating in Henan, Hubei, and Anhui. It may reflect a shift in the geographical focus of attention to the green tourism theme towards these areas. Tibet reappeared as a low-low clustering area, reaffirming its lower activity in disseminating the green tourism theme.

Fig. 14

Spatial autocorrelation and hotspot analysis of green tourism social network,