The ever increasing global population gave birth to a number of challenges that human beings are facing these days. Sustainable food development is one of them. This increase in population put a huge challenge on agriculture sector to produce more. Crops are the main source of food. Monitoring crops need accurate and real time based data that can help in planning and management. As the crops exhibit a unique character of growth that changes with time therefore, the data need to be updated on a regular time periods. Using the old fashioned methods to collect these data on regional scale is not enough to cope with these challenges. Remote sensing provides the best alternative to collect such data on regional scales that can help in crop monitoring. The advancement in the field of remote sensing data acquisition and processing imageries helped to implement different techniques for crop monitoring on large areas.　　Over the past few decades, different techniques were developed focusing on a range of crop monitoring practices using remote sensing data. With the passage of time the introduction of new sensors to remote sensing platforms provide the information to the most specific processes of crops thus helping researcher in this field. The Introduction of Moderate Resolution Imaging Spectroradiometer (MODIS) satellite made a real breakthrough especially in the studies of corpses it provides vegetation indices (VI) with a range from low to moderate resolutions both spatially and temporally, covering the whole globe. In this research MODIS data and modeling approaches are used in the following way.　　In the first step of research capabilities of MODIS EVI were tested by detecting and mapping the changes within cropland in the area of Xinjiang in 10 years from 2004 to 2013 mostly contributed due to the change in the agriculture policy of the country. Thus MODIS data is used to map the impacts of China new agriculture policy on croplands area. The comparison of remote sensing extracted results with field surveys shows that remote sensing data detect the croplands over the whole of the study area which is difficult for the field surveys to acquire such data on such large areas.　　In the second step a series of classification procedures were developed to classify major crops using subsequent/historical field reference data along with MODIS time series Enhanced Vegetation Index (EVI).These procedures are aimed to check the reliability of time series MODIS EVI 250m in crop discrimination on regional scale while using field data of other years. The use of time series MODIS EVI and subsequent field data for crop classification shows that the Artificial Antibody Network (ABNet) classifier classified the crops with high accuracy. While using historical field reference data for crop classification in single year and multiyear classes the results suggest that the classification accuracy increase in two ways, (a) when the field reference data from maximum number of years is used.(b) When the field reference data of the year nearest to the classification year is used. This step aimed to decrease the load of troublesome field data collection and make the process of crop monitoring more robust and fast to help take decisions on time.　　In the third step of the research, modeling approach is adopted to estimate the yield of the cotton crop in the KUCHI county of Xinjiang. The DSSAT-CSM model is used to simulate the growth dynamics of the cotton crop based on a bulk of weather, soil and crop management data which were used as proxies for yield estimation. In the last part of the research a proper discussion has-been made on the use of remote sensing and modeling approaches in the crop monitoring and points are concluded and recommendations for future are given.