Although crevasse splays are a common constituent of many fluvial and fluvio-deltaic systems, they remain less well understood than the channel deposits in those settings, especially with respect to controls on their occurrence, distribution, and geometry. The current study aims to redress this balance and investigate controls on 1) splay formation and occurrence and 2) splay size and geometry. The study has used Google Earth-based satellite imagery to examine crevasse-splay deposits from eight modern fluvial systems. A total of 1556 crevasse splays were identified using imagery from 1984 to 2020. Most of the splays (c. 70%) occur on the outer sinuous river bank with offtake angles ranging from 10° to 140° (mean 75°) to the channel flow direction. Three different types of splays have been identified: i) single crevasse splays, ii) laterally amalgamated crevasse splays, and iii) crevasse-splay complexes. The areal extent of splay bodies varies widely and ranges from less than 1 km2 up to 221 km2. The single crevasse splays are the primary and smallest form of splay, with an average area of 0.61 km2. Compensational stacking or progradation significantly increase the splay area and form laterally amalgamated splays and splay complexes, respectively. The average areal extent of laterally amalgamated splays is 1.33 km2, and of splay complexes, 39 km2.
The climate, discharge, floodplain morphology, vegetation, trunk channel slope, sinuosity, and sediment load primarily control the occurrence, geometry, and dimensions of crevasse splays. Results demonstrate that sparse or no floodplain vegetation favors the formation of elongated tongue-shaped crevasse splays while densely vegetated floodplains produce more lobate splays. The highest splay frequency occurs in systems where the river experiences sudden high magnitude variation in discharge, has a low cross-sectional area, and noncohesive bank materials. Larger splay size is correlated with lower river slope angles and higher sinuosity, discharge, and floodplain relief. Channel size has little influence on the extent of splays. This work suggests that autogenic factors such as trunk-channel slope and sinuosity are more influential in arid–semiarid settings while allogenic factors such as discharge are important in temperate–equatorial settings.