Iron resonance fluorescence lidar is an important means of measuring temperature and wind field in the region from the top of the intermediate layer to the bottom of the thermosphere. However, due to the presence of strong sky background noise during the daytime, the resonance fluorescence signals of iron atoms are submerged, making it impossible to detect temperature and wind field in the middle and upper atmosphere. In order to solve this problem, this paper developed a tunable narrowband filter and proposed a new filtering method of connecting two Fabry-Perot etalons with air gap in series and combining them with a narrowband interference filter. Then the filter and the method were applied in iron resonance fluorescence Doppler lidar, achieving temperature and wind field detection in the region from the top of the intermediate layer to the bottom of the thermosphere during daytime. Firstly, by defining a filter performance evaluation function, the filtering performance of the filter was quantitatively analyzed, and parameters such as bandwidth and free spectral range of the two etalons were determined. Furthermore, a control system based on PID algorithm was designed to achieve precise control of the pressure in etalon chamber. Then, the design specifications of the filter were tested through simulation. And finally, field experiments demonstrated the filtering effectiveness of the filter in the iron resonance fluorescence lidar system, achieving a signal-to-noise ratio exceeding 136 and enabling wind speed and temperature detection in the iron layer region.