As a new type of laser detection radar, Brillouin lidar can be used for monitoring seawater parameters. In view of the disadvantage of the existing Brillouin lidar inversion models that have smaller inversion accuracy and parameter range, the new inversion model for temperature and salinity is proposed by using the least squares method with frequency shift and linewidth as independent variables. This model improves the inversion accuracy and inversion range of temperature and salinity, and the inversion model is analyzed in terms of sensitivity, error and uncertainty. The results show that the Brillouin frequency shift and linewidth can be used to achieve simultaneous monitoring of ocean temperature and salinity. In the temperature inversion model, the effect of Brillouin linewidth on temperature is greater than that of frequency shift, and the temperature inversion range of the new model is extended from 15 ℃≤T≤30 ℃ to 10 ℃≤T≤30 ℃ with a maximum error of 0.17 ℃, an average error of 0.07 ℃ and an average relative error is 0.27%, which is 63.16% better than the accuracy of the existing model. In the salinity inversion model, the effect of Brillouin frequency shift on salinity is greater than that of linewidth, and the inversion range of the model is expanded from 30‰≤S≤35‰ to 15‰≤S≤35‰, with a maximum error of 0.20‰, an average error of 0.09‰ and an average relative error is 0.29%, which is 77.50% better than the accuracy of the existing model. The research results in this paper have guiding significance for Brillouin lidar to detect seawater temperature and salinity.