关键词: 红树林, 潮沟网络, 形态特征, 排水效率, Sentinel-2影像

Abstract: Tidal creek networks are fundamental geomorphic units that regulate hydrological connectivity and structural stability in mangrove ecosystems. Their spatial distribution and morphological characteristics affect drainage efficiency and vegetation patterning. Based on Sentinel-2 multispectral imagery, we integrated four widely adopted water indices (i.e., the normalized difference water index (NDWI), the modified normalized difference water index (mNDWI), automated water extraction index with shadows (AWEI_sh), and automated water extraction index without shadows (AWEI_nsh)), with a modified fuzzy C means clustering algorithm (MFCM) and an adaptive thresholding segmentation approach to delineate the mangrove tidal creek network in the Sundarbans mangrove in Bangladesh. Following visual interpretation and correction, key morphological parameters were extracted, including creek order, count, length, bifurcation ratio, density, and sinuosity. The overmarsh path length (OPL) was calculated to quantitatively characterize the drainage efficiency of the creek network in the study area. The results showed that: (1) All four water indices effectively identified mangrove tidal creeks, with AWEI_sh yielding the highest extraction accuracy. (2) Tidal creeks developed up to the 5th order, and the number of creeks decreased exponentially with increasing creek order (R²>0.99). (3) Both morphological characteristics and drainage efficiency exhibited spatial heterogeneity. Region III had the greatest number of creeks, the highest density and bifurcation ratio, but the lowest mean length and sinuosity. Region I showed the inverse pattern, with the fewest creek numbers but a greater number of high-order (fifth-order) creeks than Region III. Overall drainage efficiency ranked as follows: Region III > Region II > Region I. (4) OPL was negatively correlated with creek density, bifurcation ratio, and frequency (r=-0.86, -0.77, and -0.74, respectively, P<0.001), and positively correlated with the average creek length (r=0.60, P>0.05). This study contributes to understanding the spatial differentiation patterns of morphological characteristics and drainage efficiency in mangrove wetland tidal creek networks, providing a theoretical basis for mangrove wetland conservation and ecological restoration.

Key words: mangrove, tidal creek network, morphological characteristic, drainage efficiency, Sentinel-2 imagery