Abstract: To enhance the risk management of wind farms under fluctuating wind resources, this study uses hourly wind speed observations from wind measurement towers at two wind farms in Hebei Province as the reference to construct two types of wind power weather indices—Index Ⅰ (monthly mean wind speed) and Index Ⅱ (monthly available power generation)—based on hourly wind speed data from meteorological stations, MERRA-2, and ERA-5. Over the period from January 2022 to March 2023, basis risk analysis, pure premium rate calculation, and payout simulation are conducted. It is found that indices derived from meteorological station data exhibit the lowest basis risk across both wind farms. For tower 1 in particular, Index Ⅰ and Index Ⅱ yield correlation coefficients of 0.94 and 0.93, and basis risk volatilities of 0.042 and 0.047 GWh, respectively—significantly outperforming indices based on reanalysis datasets. Moreover, Index Ⅱ consistently shows a higher payout-to-premium ratio than Index Ⅰ, indicating stronger capacity to identify and respond to low-wind events. Under the P75 trigger level, annual pure premium rates range from 7.11% to 8.00%, with more frequent and higher payouts; under the P90 level, rates fall to 2.80%—3.05%, with reduced coverage effectiveness. The wind power weather index insurance framework developed in this study demonstrates the feasibility and applicability of designing such insurance based on wind speed data, offering robust data and methodological support for site-level weather index insurance product development.