Monitoring growth condition of spring maize in Northeast China using a process-based model

Early and accurate assessment of the growth condition of spring maize, a major crop in China, is important for the national food security. This study used a process-based Remote-Sensing-Photosynthesis-Yield Estimation for Crops (RS-P-YEC) model, driven by satellite-derived leaf area index and ground-based meteorological observations, to simulate net primary productivity (NPP) of spring maize in Northeast China from the first ten-day (FTD) of May to the second ten-day (STD) of August during 2001–2014. The growth condition of spring maize in 2014 in Northeast China was monitored and evaluated spatially and temporally by comparison with 5- and 13-year averages, as well as 2009 and 2013. Results showed that NPP simulated by the RS-P-YEC model, with consideration of multi-scattered radiation inside the crop canopy, could reveal the growth condition of spring maize more reasonably than the Boreal Ecosystem Productivity Simulator. Moreover, NPP outperformed other commonly used vegetation indices (e.g., Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI)) for monitoring and evaluating the growth condition of spring maize. Compared with the 5- and 13-year averages, the growth condition of spring maize in 2014 was worse before the STD of June and after the FTD of August, and it was better from the third ten-day (TTD) of June to the TTD of July across Northeast China. Spatially, regions with slightly worse and worse growth conditions in the STD of August 2014 were concentrated mainly in central Northeast China, and they accounted for about half of the production area of spring maize in Northeast China. This study confirms that NPP is a good indicator for monitoring and evaluating growth condition because of its capacity to reflect the physiological characteristics of crops. Meanwhile, the RS-P-YEC model, driven by remote sensing and ground-based meteorological data, is effective for monitoring crop growth condition over large areas in a near real time.