Methods and algorithms of data assimilation and its application in agriculture

Data assimilation is an advanced and innovative set of techniques for variable estimation, especially in agriculture research in recent years, which integrates not only remote sensing data products, other measurements, but also land dynamic models. It can provide more abundant and precise data set, and extend data´s spatiotemporal scale. In this paper, firstly composition and assimilation methods of land data assimilation were briefly introduced, secondly two main data assimilation algorithms such as sequential based and cost function based were mainly described, thirdly some cases of data assimilation applications in agriculture were listed. And fourthly, the limits of data assimilation in agriculture were analyzed, for instance, parameters and driven data setting, algorithm optimization, crop model perfection and remote sensing quantification. These questions may be resolved with more sources of remote sensing data, assimilation theory, crop model and development of quantitative remote sensing. Finally, the prospects of assimilation application in agriculture were analyzed.