Abstract The United States of America ranked first in maize export and second in soybean export in the world. Accurate and timely data and information on maize and soybean production in the Contiguous United States (CONUS) are important for food security at the regional and global scales. In this study, we firstly compare the maize and soybean planted area from cropland data layer (CDL) with NASS area statistics over the CONUS during 2008-2018, and evaluate the interannual changes of planted and harvested area based on the two datasets. Secondly, we investigate the relationship between grain production and gross primary production (GPP) simulated by Vegetation Photosynthesis Model (VPM) at national and county scales. Finally, we evaluate the linear regression models between grain production and cumulated GPPVPM over time at 8-day resolution. We found strong spatial-temporal consistency between CDL and NASS datasets in maize and soybean planted areas. Maize and soybean planted areas increased by mid-2010s, largely driven by markets and international trade. Severe summer drought in 2012 had little impact on soybean planted and harvested area and maize planted area, but substantially reduced maize harvested area. and grain production. Annual county-level GPPVPM had strong linear relationship with NASS grain production for maize and soybean. The Harvest Index, defined as the ratio between grain production and GPPVPM (HIGPP_VPM), ranged from 0.25 (2012) to 0.36 for maize and from 0.13 to 0.15 for soybean. The linear regression models between grain production and cumulated GPPVPM (GPPVPM_CUM) over time at 8-day resolution showed that by the end of July, GPPVPM_CUM accounted for ~90% of variance in maize and soybean grain production, which was approximately two months before farmers started to harvest. This study clearly shows that VPM and GPPVPM data are useful for monitoring and in-season forecasting of maize and soybean grain production in the CONUS.

中文翻译：

---

2008-2018 年美国本土玉米和大豆种植面积、收获面积、初级生产总值和粮食产量的时空动态

摘要 美国玉米出口量居世界第一，大豆出口量居世界第二。美国本土 (CONUS) 玉米和大豆生产的准确及时数据和信息对于区域和全球范围的粮食安全非常重要。在本研究中，我们首先将 2008-2018 年 CONUS 的农田数据层 (CDL) 中的玉米和大豆种植面积与 NASS 面积统计数据进行比较，并基于这两个数据集评估种植和收获面积的年际变化。其次，我们在国家和县级范围内研究了植被光合作用模型 (VPM) 模拟的粮食产量与初级生产总值 (GPP) 之间的关系。最后，我们以 8 天的分辨率评估了谷物产量和累积 GPPVPM 之间随时间推移的线性回归模型。我们发现玉米和大豆种植区的 CDL 和 NASS 数据集之间具有很强的时空一致性。玉米和大豆种植面积在 2010 年代中期增加，主要受市场和国际贸易的推动。2012年夏季严重干旱对大豆种植收获面积和玉米种植面积影响不大，但玉米收获面积大幅减少。和粮食生产。年度县级 GPPVPM 与 NASS 玉米和大豆的粮食产量呈强线性关系。收获指数，定义为谷物产量与 GPPVPM (HIGPP_VPM) 之间的比率，玉米的范围为 0.25 (2012) 至 0.36，大豆的范围为 0.13 至 0.15。粮食产量与 8 天分辨率累积 GPPVPM（GPPVPM_CUM）之间的线性回归模型显示，到 7 月底，GPPVPM_CUM 占玉米和大豆谷物产量变化的约 90%，这比农民开始收获前大约两个月。本研究清楚地表明，VPM 和 GPPVPM 数据对于 CONUS 玉米和大豆粮食生产的监测和季节预测非常有用。