Variations in global zonal wind from 18 to 100 km due to solar 1 activity, and QBO, ENSO during 2002–2019

23 Variations of global wind are important in changing the atmospheric structure and circulation, in coupling of 24 atmospheric layers, in influencing the wave propagations. Due to the difficulty of directly measuring zonal wind from the 25 stratosphere to the lower thermosphere, we derived a global balance wind (BU) dataset from 50°S to 50°N and during 2002– 26 2019 using the gradient wind theory and SABER temperatures and modified by meteor radar observations at the equator. 27 The dataset captures the main feature of global monthly mean zonal wind and can be used to study the variations (i.e., annual, 28 semi-annual, ter-annual, and linear) of zonal wind and the responses of zonal wind to QBO (quasi-biennial oscillation), 29 ENSO (El Niño/Southern Oscillation), and solar activity. Same procedure is performed on the MERRA2 zonal wind (MerU) 30 to validate BU and its responses below 70 km. The annual, semi-annual, ter-annual oscillations of BU and MerU have 31 similar amplitudes and phases. The semi-annual oscillation of BU has peaks around 80 km, which are stronger in the 32 southern tropical region and coincide with previous satellite observations. As the increasing of the values representing QBO 33 wind (short for QBO wind), both values of representing BU and MerU (short for BU and MerU) change from increasing to 34 decreasing with the increasing height and extend from the equator to higher latitudes. Both BU and MerU increase with the 35 increasing of the values of MEI (an indicator of ENSO, short for MEI) and decrease with increasing F10.7 (an indicator of 36 solar activity) in the southern stratospheric polar jet region below 70 km. The responses of winds to ENSO and F10.7 exhibit 37 hemispheric asymmetry and are more significant in the southern polar jet region. While above 70 km, BU increases with the 38 increasing of MEI and F10.7. The negative linear changes of BU at 50°N are absent in MerU during October–January. The 39 discussions on the possible influences of the temporal intervals and sudden stratospheric warmings (SSWs) on the variations 40 and responses of BU illustrate that: (1) the seasonal variations and the responses to QBO are almost independent on the 41 temporal intervals selected; (2) the responses to ENSO and F10.7 are robust but slightly depend on the temporal intervals; (3) 42 the linear changes of both BU and MerU depend strongly on the temporal intervals; (4) SSWs affect the magnitudes but do 43 not affect the hemispheric asymmetry of the variations and responses of BU at least in the monthly mean sense. The 44 variations and responses of global zonal wind to various factors are based on BU, which is derived from observations, and 45 thus provide a good complementary to model studies and ground-based observations.