A comparison of retrieval approaches for estimating the seasonal dynamics of rice leaf area index from simulated Sentinel-2 data

O. A. Adeluyi; A. Harris; G. Clay; T. Foster

doi:10.1117/12.2533304

21 October 2019 A comparison of retrieval approaches for estimating the seasonal dynamics of rice leaf area index from simulated Sentinel-2 data

O. A. Adeluyi, A. Harris, G. Clay, T. Foster

Author Affiliations +

Proceedings Volume 11149, Remote Sensing for Agriculture, Ecosystems, and Hydrology XXI; 111490F (2019) https://doi.org/10.1117/12.2533304
Event: SPIE Remote Sensing, 2019, Strasbourg, France

Abstract

Leaf Area Index (LAI) is a key variable for the monitoring the ability of crops to intercept solar energy for biomass production. In addition, LAI has the high seasonal variability and differ among phenophases. Therefore, accurately monitoring LAI at the vegetative, reproductive and ripening stage is central for improving yield. The aim of the study is to identify the performance of different retrieval methods on the estimation of LAI at different phenophases of rice from sentinel 2 simulated bands. To achieve this, the research seeks to answer the following research objectives (i) compare the performance of piecewise model based on the specific phenophases with single models generated from the entire active season, (ii) to determine whether phenophase models are advantageous for LAI estimation in rice. The retrieval models were developed and tested on proximal hyperspectral bands, with focus on the sentinel 2 spectral bands. The machine learning regression models (MLRAs) presented the highest retrieval accuracy during the entire growing season (R2=0.65-0.74). The hybrid model’s retrieval performance was similar to vegetation indices but with much higher errors during the entire growing season. The phenophases estimation of LAI saw a decline in the overall retrieval performance of MLRA, hybrid and VI models. The models performed much better during the vegetative stage compared to the reproductive and ripening stages. The study shows MLRA as the best model for estimating LAI during the entire growing and vegetative stages of rice growth. The hybrid models were only suited for the entire growing season (R2>0.5) but generally low when estimating for different phenohases. Finally, VI models were identified to be the best for estimating LAI during the ripening stages of irrigated rice.

Conference Presentation

Citation Download Citation

O. A. Adeluyi, A. Harris, G. Clay, and T. Foster "A comparison of retrieval approaches for estimating the seasonal dynamics of rice leaf area index from simulated Sentinel-2 data", Proc. SPIE 11149, Remote Sensing for Agriculture, Ecosystems, and Hydrology XXI, 111490F (21 October 2019); https://doi.org/10.1117/12.2533304

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available