Introducing a Farmer-Assisted Biomass Estimation (FABE) model using satellite images

• Published in: Advances in space research Vol. 66; no. 7; pp. 1522 - 1536
• Main Authors: Hejazi, S. Abbas, Mobasheri, Mohammad Reza
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.10.2020
• Subjects: Biomass; Landsat; Satellite image; Wheat; Satellite image; Biomass; Wheat; Landsat
• ISSN: 0273-1177; 1879-1948
• DOI: 10.1016/j.asr.2020.06.009

•More accurate estimation of wheat biomass in regional to global scale.•Ability of this model to work even when the sky is overcast.•Enhancement of estimation using farmers-supplied auxiliary data. Assessment of the wheat biomass for yield prediction in regional to global scales is of particular importance, especially in food security. So far, many studies have been conducted in this subject where almost in all of them either satellite data or climatological data have been deployed. In the present study, first, some commonly used models including empirical and conventional Monteith models for wheat biomass estimate on have been evaluated where the results were found to be not quite satisfactory. Then a model named FABE, based on the farmer-provided auxiliary data along with the satellite and climatological data has been introduced. The uncertainty of the FABE model in estimating biomass was 193 (gm−2) with relative RMSE of about 0.31. This shows significant improvement compared to models such as Monteith (RMSE = 224 (gm−2) and relative RMSE = 0.36). It is shown that the biomass prediction improves if it benefits from farmer-supplied auxiliary data including wheat farms location, watering system, and digital photographs to fill up the gaps due to the satellite missing pixels/images. Accurate biomass estimation may improve the prediction of the yield.