<p>Agricultural irrigation in India is often based on subjective and regionally variable practices, leading to poor representation in land surface models (LSMs). Data assimilation (DA) is an effective method to integrate soil moisture (SM) observations into model predictions, to produce an observation based, spatially distributed SM estimates over Indian domian. Previous studies have employed DA to incorporate unmodeled processes, such as irrigation, into model SM estimates. However, these studies had limited success, primarily due to ineffective bias correction methods. To overcome the above limitation, the present study employed and tested an anomaly-based bias correction method prior to DA, for the first time over the Indian domain. The present study assimilated Soil Moisture Active Passive (SMAP) satellite retrievals into the Noah LSM, utilizing Global Data Assimilation System (GDAS) atmospheric forcings data and precipitation data from the following three data sets such as, TRMM, GDAS, and IMERG-GPM over the Indian domain. DA with the anomaly correction method performs better, as compared to DA with the cumulative distribution function (CDF) matching method, particularly during the winter and pre-monsoon seasons. RMSE values for DA with the anomaly correction are lower during pre-monsoon and winter seasons for all the three different precipitation-forced SM estimates. This improvement is attributed to the significant irrigation over India during pre-monsoon and winter seasons. The comparison with the Global Map of Irrigation Area (GMIA) showed that improvements of GDAS forced DA with anomaly correction method are significant over highly irrigated regions. This study highlights the effectiveness of assimilating SMAP data with anomaly-based bias correction in improving soil moisture estimates from the Noah LSM over India. The approach not only enhances model accuracy but also helps reveal irrigation signals, particularly during high-irrigation seasons.</p>