A simplified approach to satellite-based monitoring system of sugarcane plantation to manage yield decline at Wonji-Shoa Sugar Estate, central Ethiopia

Heliyon · 4 authors, 4 centres

AI SUMMARY

FIDELITY 100%

This summary was generated by AI from a single paper. It has not been reviewed by a clinician and is not clinical advice. Verify against the source before acting on it.

AGENT 01 · STRUCTURED SUMMARY

TRACE TO SOURCE

“

TL;DR

This study evaluated ten vegetation indexes calculated by the user-friendly Landviewer satellite platform for monitoring sugarcane plantations in central Ethiopia. The indexes significantly followed the sigmoid growth curve of sugarcane and showed strong correlations with yield components, fractional green canopy cover, and final yield, with the Normalized Difference Phenology Index (NDPI), Normalized Difference Vegetation Index (NDVI), and Soil-Adjusted Vegetation Index (SAVI) being the most effective. These findings provide a practical, low-cost tool for developing countries to combat yield decline through timely crop monitoring.

Full summary

4,184 CHARS

**Background:** Sugarcane yields at Wonji-Shoa Sugarcane Plantation (WSSP) in central Ethiopia have declined by 48% over the past 70 years. Conventional field monitoring is inefficient due to dense canopy, safety hazards (e.g., snakes, corneal injuries from leaves), and high costs. Satellite-based monitoring offers a safer, faster alternative, but traditional remote sensing requires advanced skills and computing resources unavailable in many African countries. The Landviewer platform (EOS Data Analytics) provides on-the-fly satellite image processing via a browser, eliminating these barriers. This study aimed to evaluate ten Landviewer-calculated vegetation indexes (LCVIs) for monitoring sugarcane and identify the most effective indexes.

**Methods:** The study was conducted at WSSP during the 2020/2021 and 2021/2022 cropping seasons. Sentinel-2A satellite imagery (10–60 m spatial resolution, 5-day revisit) was used. Ten VIs were evaluated: Enhanced Normalized Difference Vegetation Index (ENDVI), Green Differential Vegetation Index (GDVI), Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), Enhanced Vegetation Index (EVI), Normalized Difference Phenology Index (NDPI), Normalized Difference RedEdge (NDRE), Red-edge Chlorophyll Index (ReCl), Ratio Vegetation Index (RVI), and Green Normalized Difference Vegetation Index (GNDVI). Four approaches were used: (1) goodness-of-fit of sigmoid growth curves (logistic for plant cane, Gompertz for ratoon cane) over the first 10 months; (2) relationship between LCVIs and yield components (plant density, height, diameter) in three major varieties (NCo334, N14, B52298); (3) relationship between LCVIs and fractional green canopy cover (FGCC) measured via Canopeo software from smartphone images taken at 3.5 m height; (4) relationship between time-series NDVI (generated by Landviewer) and final yield from 87 ratoon fields (29 per variety). Statistical analyses included regression and Tukey's test at 95% confidence.

**Key Results:** All LCVIs significantly followed the sigmoid growth curve (p < 0.001). For plant cane, mean percent variance accounted for ranged from 83.91% (ReCl) to 95.13% (SAVI); for ratoon cane, from 80.58% (ReCl) to 93.03% (SAVI). SAVI, NDPI, NDRE, EVI, GNDVI, GDVI, and ENDVI (91–95% variability) performed significantly better than RVI and ReCl (80–89% variability). Varietal differences were significant: for plant cane, NCo334 (95%) and B52298 (94%) outperformed N14 (86%); for ratoon cane, NCo334 (93%) and N14 (91%) outperformed B52298 (85%). Yield components showed strong correlations: in NCo334, GDVI had R² = 0.90 with plant density, GNDVI had R² = 0.74 with height and R² = 0.65 with diameter; combined R² ranged from 0.70 (ENDVI) to 0.93 (GDVI). In N14, SAVI and EVI had R² = 0.88 with height; combined R² ranged from 0.72 (NDVI) to 0.92 (NDPI). In B52298, only plant density showed significant correlations (GDVI R² = 0.66, EVI R² = 0.64); combined R² was significant only for EVI (0.79). FGCC correlations: in NCo334, SAVI had R² = 0.79; in N14, SAVI had R² = 0.93; in B52298, NDVI had R² = 0.83. NDPI, NDVI, and GNDVI were significant for all three varieties. Time-series NDVI correlated significantly with yield from 6 months onward: at harvest, R² = 0.85 for NCo334, 0.81 for N14, and 0.73 for B52298. For NCo334, fields with maximum yield (134 ton/ha) had NDVI values 16–30% higher than minimum-yield fields (24.4 ton/ha) at various ages.

**Clinical Implications:** This study validates that Landviewer-calculated VIs are as effective as conventionally processed satellite indexes for monitoring sugarcane, offering a simple, user-friendly, and fast tool that can be operated by non-experts on ordinary computers or smartphones. The platform can help plantation managers detect growth anomalies early, enabling timely interventions to mitigate yield decline. The annual yield loss at WSSP is estimated at US$ 8,228,558. Limitations include focus on only three varieties, single location, and inability to identify specific stress types. Future work should evaluate additional varieties, locations, and stress detection features of Landviewer.

STUDY TYPE

other

SPECIALTY

agriculture

SUMMARISED BY

AI pipeline

FIDELITY CHECK

100% · A

ORIGINAL PAPER

PMC10432716

A simplified approach to satellite-based monitoring system of sugarcane plantation to manage yield decline at Wonji-Shoa Sugar Estate, central Ethiopia

Heliyon

READ THE FULL PAPER

JUMP TO SECTION

ABSTRACT

Abstract Drastic and continuous decline in cane yields has become a major threat to sustainable sugarcane production in Ethiopia. Among the causes for the decline are the inefficient and ineffective system of monitoring sugarcane plantations. Adopting satellite-based crop monitoring through the Landviewer platform may circumvent this problem. However, the reliability of vegetation indexes calculated by the platform is unknown and thus requires evaluation. Accordingly, we tested the accuracy of selected Landviewer Calculated Vegetation Indexes (LCVIs) on three major sugarcane varieties and two cropping types. The goodness-of-fit of the sigmoid curve to the LCVIs profile of sugarcane was evaluated. The correlations between LCVIs and yield components, LCVIs and fractional green canopy cover (FGCC), as well as the time-serious Normalized Difference Vegetation Index (NDVI) and yields, were also analysed. We found that the goodness-of-fit of the sigmoid curve was significant ( p < 0.001), with 84%–95% accuracy in all the indexes. The majority of LCVIs showed significant ( p < 0.05) relationships with yield components and FGCC. The time-series NDVI also demonstrated a significant relationship with cane yield (R 2 = 0.73–0.85) at the age of 10 months and above. The accuracy level of LCVIs varies with varieties and crop types, but the Normalized Difference Phenology Index (NDPI), Soil Adjusted Vegetation Index (SAVI), and NDVI were identified as the most consistent and effective LCVIs for sugarcane monitoring. Therefore, the accuracy of LCVIs was dependable and can be used effectively in monitoring sugarcane plantations to tackle the problem of continuous decline in the yield of the crop. Keywords: Canopy cover, Correlation, Landviewer, Platform, Sigmoid curve, Vegetation indexes, Yield components

CITATION EXPORT

Alemayehu Dengia, Nigussae Dechassa, Lemma Wogi et al.. A simplified approach to satellite-based monitoring system of sugarcane plantation to manage yield decline at Wonji-Shoa Sugar Estate, central Ethiopia. Heliyon. (2023). https://doi.org/10.1016/j.heliyon.2023.e18982

AUTHORS

4 · 4 CENTRES

Alemayehu Dengia, Nigussae Dechassa, Lemma Wogi, Berhanu Amsalu

aEhiopian Sugar Industry Group, Research and Training, P. O. Box,15, Wonji, Adama, Ethiopia
bHaramaya University, College of Agriculture and Environmental Sciences, Africa Center of Excellence for Climate Smart Agriculture and Biodiversity Conservation, P. O. Box 138, Dire Dawa, Ethiopia
cHaramaya University, School of Natural Resources Management and Environmental Sciences, P. O. Box 138, Dire Dawa, Ethiopia
dEthiopian Institute of Agricultural Research (EIAR), Melkassa Agricultural Research Centre, P.O. Box 436, Adama, Ethiopia