Home» Research» Publication» IARRP scientists propose a novel remote sensing data assimilation algorithm for regional yield estimation

IARRP scientists propose a novel remote sensing data assimilation algorithm for regional yield estimation

IARRP | Updated: 2021-01-21

Scientists from the Innovation Team of Smart Agriculture of the Institute of Agricultural Resources and Regional Planning (IARRP) of the Chinese Academy of Agricultural Sciences (CAAS), have made important progress in the research of remote sensing data assimilation technology for regional crop yield estimation by proposing the WOFOST model and a novel VW-4DEnSRF assimilation algorithm, constructing a crop yield estimation assimilation system, and conducting high-precision regional crop yield simulation.

The research results were published online in the top international academic journal Remote Sensing of Environment (impact factor 9.085) in an article titled “Regional winter wheat yield estimation based on the WOFOST model and a novel VW-4DEnSRF assimilation algorithm”.

The general empirical statistical method for crop yield estimation by remote sensing cannot quantitatively describe the crop growth process. It is important to give full play to the advantages of a crop growth model, such as strong mechanism, time continuity, and spatial continuity of remote sensing data, and to further improve the quantitative simulation and estimation accuracy of regional crop yield based on the remote sensing information and growth model assimilation.

Since there are difficulties in obtaining and correcting large-scale crop model parameters, and the commonly used Kalman filter assimilation algorithm has shortcomings such as non-convergence and singular values, and the four-dimensional variational algorithm has defects such as a fixed value of background error, the constructed crop yield estimation assimilation system cannot fully meet the high-precision simulation requirements for temporal and spatial variability of crop growth.

In this study, a novel EnSRF assimilation algorithm based on a variable time window and four-dimensional extension (VW-4DEnSRF) was proposed. Based on the WOFOST crop model and the proposed VW-4DEnSRF algorithm, a crop yield assimilation system was successfully constructed after parameter sensitivity analysis and parameter calibration of the crop model.

Taking Hengshui city of Hebei province as the study area and winter wheat as the research crop, the researchers used the leaf area index information retrieved from GF-1 and HJ-1 domestic satellite data as external remote sensing assimilation data, and realized the quantitative simulation and estimation of regional winter wheat yield with the constructed crop yield estimation assimilation system under the optimal scale grid.

Through comparison with the field-measured yield data and official statistical yield data at the county level, the researchers found that the crop yield assimilation system based on the WOFOST model and proposed VW-4DEnSRF algorithm performed well at both the single-point level and regional level, which proved that the proposed algorithm was feasible and effective at simulating crop yield over a large area.

The VW-4DEnSRF assimilation algorithm and the assimilation system proposed in this study are a useful supplement to the existing international remote sensing assimilation algorithms and assimilation systems, and improve the accuracy of crop yield simulation estimation by remote sensing assimilation algorithms. It also has important application value for the future development of large-scale crop yield simulation forecasts, regional crop growth monitoring and evaluation, and national food security.

Figure 1 A novel VW-4DEnSRF assimilation algorithm framework and regional winter wheat yield quantitative simulation results

This research was funded by the National Natural Science Foundation of China for innovation research group, the Young Talent Support Project of the China Association for Science and Technology, and the Scientific and Technological Innovation Project of the Chinese Academy of Agricultural Sciences. Dr. Wu Shangrong is the first author, and research fellows Ren Jianqiang and Chen Zhongxin are the corresponding authors.

(1)  https://www.sciencedirect.com/science/article/pii/S0034425720306490

(2)  https://authors.elsevier.com/a/1cQBe7qzSr7iw

Wu Shangrong, Yang Peng, Ren Jianqiang*, Chen Zhongxin*, Li He. Regional winter wheat yield estimation based on the WOFOST model and a novel VW-4DEnSRF assimilation algorithm [J]. Remote Sensing of Environment, 2021, 255(15):112276.