This research is dedicated to the study of the feasibility of surface wind downscaling from 925 or 850 hPa winds according to synoptic class, season and hour. Two aspects are examined: low tropospheric wind veering and wind speed correlation and verification of the ERA-Interim analysis wind by comparison to radiosonde data at Beit Dagan, a station on the Israeli coast. Relatively small (< 60°) cross angles between the 1000 hPa wind vector and the 925 hPa or 850 hPa wind vector at 12Z and high correlation (0.6-0.8) between the wind speed at the two levels were found only under winter lows. Relatively small cross angles and small wind speed correlation were found under highs to the west and Persian troughs. This suggests the possibility to directly infer surface wind/wind direction from low troposhere wind under these cases. The verification of ERA-Interim analysis in comparison with radiosonde data has shown good prediction of wind direction at 12Z at 1000, 925 and 850 hPa levels (RMSE 20°-60°) and lower prediction quality at 1000 hPa at 0Z (RMSE 60°-90°). The analysis under-predicts the wind speed, especially at 1000 hPa. The wind speed RMSE is 1-2 m/s, except for winter lows with 2-3 m/s RMSE at 0Z, 12Z at all levels. In conclusion, inference of surface wind may be possible at 12Z from 925 or 825 hPa winds under winter lows. Inference of wind direction from 925 hPa winds may be possible under highs to the west and Persian troughs. Wind speed should be inferred by interpolation, according to historical data of measurements or high resolution model.