Scholarlink

Purpose <p>There is growing interest in using remote sensing tools to monitor crop status and develop in-season management strategies for optimizing rice productivity. However, the degree to which a vegetation index obtained from satellite imagery at the panicle initiation (PI) growth stage is associated with final yield remains underexplored. The objectives of this study were to (i) evaluate how Normalized Difference Red Edge Index (NDRE) at PI influences the likelihood of achieving top or bottom rice yields across commercial fields and (ii) explore key management practices that differentiate these fields.</p> Methods <p>We analyzed data from 141 commercial rice fields in Uruguay across three growing seasons (2018–2020), covering approximately 18,000&#xa0;ha. Spectral data were obtained from Sentinel-2 satellite imagery processed in Google Earth Engine, using near-infrared (864&#xa0;nm) and red-edge (704&#xa0;nm) bands to calculate NDRE at a 20&#xa0;m spatial resolution. NDRE values at PI were classified into terciles (high, medium, and low), and yield quintiles were used to define top and bottom yield classes. We assessed the frequency of fields transitioning between NDRE and yield classes and potential management factors associated with these yield trajectories.</p> Results <p>Fields with high NDRE at PI had a 30% probability of achieving top yields (&gt; 80th percentile) and only a 12% probability of falling into the bottom yield class (&lt; 20th percentile) (p &lt; 0.05). However, high NDRE alone was not sufficient, as 12% of fields with high NDRE still ended up with yields in the bottom class. A key distinguishing factor was the fractioning of N applications: fields that maintained high yields ("stay high") received 40% more N at PI, which was associated with a yield advantage of 1,615&#xa0;kg&#xa0;ha<sup>−1</sup> compared to fields that declined ("going down", 9,224&#xa0;kg&#xa0;ha<sup>−1</sup>), despite similar total N rates. Sustained irrigation and lower weed pressure were also associated with higher yields.</p> Conclusions <p>NDRE at PI is a useful indicator of potential yield outcomes but does not fully determine final yield. Management decisions, particularly N fertilization fractioning and irrigation practices, were associated with field trajectories. Findings from this study support the use of satellite-derived vegetation indices at PI as a valuable tool for identifying high-yield potential fields and guiding management strategies to sustain productivity and optimize resource use.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Does a healthy rice crop at panicle initiation guarantee high yields? combining field-level data and satellite remote sensing

  • 期刊论文

  • 发表日期：

    2026年4月30日

查看全文
  • Ignacio Macedo,
  • Alvaro Roel,
  • Cameron M. Pittelkow

摘要

Purpose

There is growing interest in using remote sensing tools to monitor crop status and develop in-season management strategies for optimizing rice productivity. However, the degree to which a vegetation index obtained from satellite imagery at the panicle initiation (PI) growth stage is associated with final yield remains underexplored. The objectives of this study were to (i) evaluate how Normalized Difference Red Edge Index (NDRE) at PI influences the likelihood of achieving top or bottom rice yields across commercial fields and (ii) explore key management practices that differentiate these fields.

Methods

We analyzed data from 141 commercial rice fields in Uruguay across three growing seasons (2018–2020), covering approximately 18,000 ha. Spectral data were obtained from Sentinel-2 satellite imagery processed in Google Earth Engine, using near-infrared (864 nm) and red-edge (704 nm) bands to calculate NDRE at a 20 m spatial resolution. NDRE values at PI were classified into terciles (high, medium, and low), and yield quintiles were used to define top and bottom yield classes. We assessed the frequency of fields transitioning between NDRE and yield classes and potential management factors associated with these yield trajectories.

Results

Fields with high NDRE at PI had a 30% probability of achieving top yields (> 80th percentile) and only a 12% probability of falling into the bottom yield class (< 20th percentile) (p < 0.05). However, high NDRE alone was not sufficient, as 12% of fields with high NDRE still ended up with yields in the bottom class. A key distinguishing factor was the fractioning of N applications: fields that maintained high yields ("stay high") received 40% more N at PI, which was associated with a yield advantage of 1,615 kg ha−1 compared to fields that declined ("going down", 9,224 kg ha−1), despite similar total N rates. Sustained irrigation and lower weed pressure were also associated with higher yields.

Conclusions

NDRE at PI is a useful indicator of potential yield outcomes but does not fully determine final yield. Management decisions, particularly N fertilization fractioning and irrigation practices, were associated with field trajectories. Findings from this study support the use of satellite-derived vegetation indices at PI as a valuable tool for identifying high-yield potential fields and guiding management strategies to sustain productivity and optimize resource use.