Abstract: Tropospheric delay is one of the main factors that affect the performance of precise point positioning (PPP), especially the accuracy of elevation direction solution. Generally, model correction and parameter estimation are used to deal with it. The accuracy of elevation direction solution is very important for gravity field recovery in airborne gravimetry. Therefore, GMF, NMF, VMF1 and VMF3 are respectively used as mapping functions, and 4 \rmm\rmm/\sqrt\rmh, 10 \rmm\rmm/\sqrt\rmh, 50 \rmm\rmm/\sqrt\rmh and 100 \rmm\rmm/\sqrt\rmh are respectively selected as random walk noise parameters to compare and analyze the influence of different tropospheric models on precise point positioning of airborne large dynamic data. The experimental results show that different tropospheric models have great influence on the positioning results in north (N) and up (U) directions, and the model difference can reach 3 mm to 4 mm; VMF1 and VMF3 models are superior to GMF and NMF models in positioning accuracy; selecting 4\;\rmm\rmm/\sqrt\rmh as random walk noise has higher positioning accuracy than the other three models.