The processes of pollen dispersal and deposition in inland lakes are essential for pollen-based paleoclimate and paleoenvironment reconstructions. However, systematic investigations assessing basin-scale airborne and fluvial influences on pollen distributions in lake-surface sediments remain limited. The implications of these processes for pollen-based reconstructions remain unclear. We use Sugan Lake, an arid endorheic basin, to examine modern pollen assemblages from airborne and lake-surface sediment samples. We evaluate spatial patterns within the lake, identify pollen sources, and assess their effects on precipitation reconstructions. In deep-water samples, pollen assemblages are dominated by regionally derived drought-tolerant taxa (Artemisia, Chenopodioideae), with hygrophilous taxa (Cyperaceae, Poaceae) and minor arboreal components (Pinaceae, Betulaceae, Juglandaceae). In contrast, shallow-water samples (<6 m) contain higher Poaceae linked to nearshore wetland inputs. No significant differences are observed between assemblages from river inflow zones and other nearshore sites, indicating minimal fluvial influence on spatial heterogeneity. Airborne samples resemble lake-surface assemblages but contain higher arboreal pollen, reflecting windborne input. Backward-trajectory modeling suggests greater contributions from northern sources during April-August and western sources during November-March. To isolate basin-derived precipitation signals, airborne influence was estimated using arboreal pollen as a tracer and removed from lake assemblages. Corrections to lake-surface pollen assemblages for airborne inputs led to better precipitation reconstructions. These results align with instrumental observations and regional arid-climate conditions. This research offers a practical framework to mitigate airborne-related bias in pollen-based reconstructions from arid lake basins, emphasizing the integration of airborne pollen monitoring and fluvial assessments to account for key inputs.