Development of Machine Learning Models to Predict Compressed Sward Height in Walloon Pastures Based on Sentinel-1, Sentinel-2 and Meteorological Data Using Multiple Data Transformations

Accurate information about the available standing biomass on pastures is critical for the adequate management of grazing and its promotion to farmers. In this paper, machine learning models are developed to predict available biomass expressed as compressed sward height (CSH) from readily accessible meteorological, optical (Sentinel-2) and radar satellite data (Sentinel-1). This study assumed that combining heterogeneous data sources, data transformations and machine learning methods would improve the robustness and the accuracy of the developed models. A total of 72,795 records of CSH with a spatial positioning, collected in 2018 and 2019, were used and aggregated according to a pixel-like pattern. The resulting dataset was split into a training one with 11,625 pixellated records and an independent validation one with 4952 pixellated records. The models were trained with a 19-fold cross-validation. A wide range of performances was observed (with mean root mean square error (RMSE) of cross-validation ranging from 22.84 mm of CSH to infinite-like values), and the four best-performing models were a cubist, a glmnet, a neural network and a random forest. These models had an RMSE of independent validation lower than 20 mm of CSH at the pixel-level. To simulate the behavior of the model in a decision support system, performances at the paddock level were also studied. These were computed according to two scenarios: either the predictions were made at a sub-parcel level and then aggregated, or the data were aggregated at the parcel level and the predictions were made for these aggregated data. The results obtained in this study were more accurate than those found in the literature concerning pasture budgeting and grassland biomass evaluation. The training of the 124 models resulting from the described framework was part of the realization of a decision support system to help farmers in their daily

[1]  E. Pebesma,et al.  Classes and Methods for Spatial Data , 2015 .

[2]  Claudia Notarnicola,et al.  Exploiting Time Series of Sentinel-1 and Sentinel-2 Imagery to Detect Meadow Phenology in Mountain Regions , 2019, Remote. Sens..

[3]  Ross N. Cuthbert,et al.  Micro-sonic sensor technology enables enhanced grass height measurement by a Rising Plate Meter , 2019, Information Processing in Agriculture.

[4]  T. McCarthy,et al.  GrassQ - A holistic precision grass measurement and analysis system to optimize pasture based livestock production , 2019, 2019 Boston, Massachusetts July 7- July 10, 2019.

[5]  Sébastien Lê,et al.  FactoMineR: An R Package for Multivariate Analysis , 2008 .

[6]  J. Thornley.  Grassland Dynamics: An Ecosystem Simulation Model , 1998 .

[7]  Pietro Ceccato,et al.  Can we Monitor Height of Native Grasslands in Uruguay with Earth Observation? , 2019, Remote. Sens..

[8]  K. Nakagami.  A method for approximate on-farm estimation of herbage mass by using two assessments per pasture , 2016 .

[9]  Cameron E. F. Clark,et al.  Original paper: Use of a pasture growth model to estimate herbage mass at a paddock scale and assist management on dairy farms , 2010 .

[10]  L. Shalloo,et al.  Increasing Grazing in Dairy Cow Milk Production Systems in Europe , 2020, Sustainability.

[11]  Josef Hakl,et al.  The use of a rising plate meter to evaluate lucerne (Medicago sativa L.) height as an important agronomic trait enabling yield estimation , 2012 .

[12]  Onisimo Mutanga,et al.  Monitoring the Spatio-Temporal Variations of C3/C4 Grass Species Using Multispectral Satellite Data , 2018, IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium.

[13]  Michael Wachendorf,et al.  Fusion of Ultrasonic and Spectral Sensor Data for Improving the Estimation of Biomass in Grasslands with Heterogeneous Sward Structure , 2017, Remote. Sens..

[14]  Georg Bareth,et al.  Estimating biomass in temperate grassland with high resolution canopy surface models from UAV-based RGB images and vegetation indices , 2019, Journal of Applied Remote Sensing.

[15]  Yu Liu,et al.  A comparative study of relevant vector machine and support vector machine in uncertainty analysis , 2013, 2013 International Conference on Quality, Reliability, Risk, Maintenance, and Safety Engineering (QR2MSE).

[16]  Alireza Taravat,et al.  Automatic Grassland Cutting Status Detection in the Context of Spatiotemporal Sentinel-1 Imagery Analysis and Artificial Neural Networks , 2019, Remote. Sens..

[17]  C. R. Eastwood,et al.  Using a ‘network of practice’ approach to match grazing decision-support system design with farmer practice , 2017 .

[18]  Matthew F. McCabe,et al.  Daily Retrieval of NDVI and LAI at 3 m Resolution via the Fusion of CubeSat, Landsat, and MODIS Data , 2018, Remote. Sens..

[19]  L. Shalloo,et al.  Review: Grass-based dairy systems, data and precision technologies. , 2018, Animal : an international journal of animal bioscience.

[20]  Matthew F. McCabe,et al.  High-Resolution NDVI from Planet's Constellation of Earth Observing Nano-Satellites: A New Data Source for Precision Agriculture , 2016, Remote. Sens..

[21]  A. González‐Rodríguez,et al.  Measuring grass yield by non-destructive methods. , 2003 .

[22]  Paul Rodríguez,et al.  A systematic literature review on the use of machine learning in precision livestock farming , 2020, Comput. Electron. Agric..

[23]  G. Bellocchi,et al.  The use of biogeochemical models to evaluate mitigation of greenhouse gas emissions from managed grasslands. , 2018, The Science of the total environment.

[24]  Emilio A. Laca,et al.  Comparison of weight estimate and rising-plate meter methods to measure herbage mass of a mountain meadow , 1989 .

[25]  Marie-Josée Cros,et al.  A biophysical dairy farm model to evaluate rotational grazing management strategies , 2003 .

[26]  R. Houborg,et al.  A Cubesat enabled Spatio-Temporal Enhancement Method (CESTEM) utilizing Planet, Landsat and MODIS data , 2018 .

[27]  D. Veselko,et al.  Evolution of Grazing Practices in Belgian Dairy Farms: Results of Two Surveys , 2019, Sustainability.

[28]  Liina Talgre,et al.  Relating Sentinel-1 Interferometric Coherence to Mowing Events on Grasslands , 2016, Remote. Sens..

[29]  D. Moot,et al.  Base temperature and thermal time requirements for germination and emergence of temperate pasture species , 2000 .

[30]  Georg Bareth,et al.  Replacing Manual Rising Plate Meter Measurements with Low-cost UAV-Derived Sward Height Data in Grasslands for Spatial Monitoring , 2018, PFG – Journal of Photogrammetry, Remote Sensing and Geoinformation Science.

[31]  Chengci Chen,et al.  Using remote sensing to estimate forage biomass and nutrient contents at different growth stages , 2018, Biomass and Bioenergy.

[32]  The use of compressed height to estimate the yield of a differently fertilized meadow , 2018 .

[33]  A. Elgersma.  Grazing increases the unsaturated fatty acid concentration of milk from grass‐fed cows: A review of the contributing factors, challenges and future perspectives , 2015 .

[34]  Brian O'Connor,et al.  Analysis of Sentinel-2 and RapidEye for Retrieval of Leaf Area Index in a Saltmarsh Using a Radiative Transfer Model , 2019, Remote. Sens..

[35]  Onisimo Mutanga,et al.  Determining Optimal New Generation Satellite Derived Metrics for Accurate C3 and C4 Grass Species Aboveground Biomass Estimation in South Africa , 2018, Remote. Sens..

[36]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[37]  S. Recous,et al.  STICS : a generic model for the simulation of crops and their water and nitrogen balances. I. Theory, and parameterization applied to wheat and corn , 1998 .

[38]  Stuart Green,et al.  Application of statistical and machine learning models for grassland yield estimation based on a hypertemporal satellite remote sensing time series , 2014, 2014 IEEE Geoscience and Remote Sensing Symposium.

[39]  H. Wickham,et al.  A Grammar of Data Manipulation , 2015 .

[40]  Aavudai Anandhi,et al.  Growing degree days – Ecosystem indicator for changing diurnal temperatures and their impact on corn growth stages in Kansas , 2016 .

[41]  P. Starks,et al.  Estimating leaf area index and aboveground biomass of grazing pastures using Sentinel-1, Sentinel-2 and Landsat images , 2019, ISPRS Journal of Photogrammetry and Remote Sensing.

[42]  J. Bailey,et al.  Improving productivity and increasing the efficiency of soil nutrient management on grassland farms in the UK and Ireland using precision agriculture technology , 2019, European Journal of Agronomy.

[43]  Luc Delaby,et al.  Development of the Moorepark St Gilles grass growth model (MoSt GG model): A predictive model for grass growth for pasture based systems , 2018, European Journal of Agronomy.

[44]  O. Balocchi,et al.  Water-Soluble Carbohydrate Recovery in Pastures of Perennial Ryegrass (Lolium perenne L.) and Pasture Brome (Bromus valdivianus Phil.) Under Two Defoliation Frequencies Determined by Thermal Time , 2020 .

[45]  Andriamasinoro Lalaina Herinaina Andriamandroso,et al.  Mapping and Monitoring of Biomass and Grazing in Pasture with an Unmanned Aerial System , 2019, Remote. Sens..

[46]  A. Verhoefa,et al.  Application of Sentinel 2 A data for pasture biomass monitoring using a physically based radiative transfer model , 2018 .

[47]  Matthew F. McCabe,et al.  A hybrid training approach for leaf area index estimation via Cubist and random forests machine-learning , 2018 .

[48]  E. Vierling,et al.  Exceptional sensitivity of Rubisco activase to thermal denaturation in vitro and in vivo. , 2001, Plant physiology.

[49]  Guangqing Chi,et al.  Applied Spatial Data Analysis with R , 2015 .

[50]  Federico Filipponi,et al.  Sentinel-1 GRD Preprocessing Workflow , 2019, Proceedings.

[51]  E. Honkavaara,et al.  Machine learning estimators for the quantity and quality of grass swards used for silage production using drone-based imaging spectrometry and photogrammetry , 2020 .

[52]  Corine Davids,et al.  Yield Estimates by a Two-Step Approach Using Hyperspectral Methods in Grasslands at High Latitudes , 2019, Remote. Sens..

[53]  Alternative Methods of Estimating Forage Height and Sward Capacitance in Pastures Can Be Cross Calibrated , 2007 .

[54]  Cawkwell,et al.  Satellite remote sensing of grasslands : from observation to management , 2016 .

[55]  Peter Beale,et al.  Spatial and Temporal Pasture Biomass Estimation Integrating Electronic Plate Meter, Planet CubeSats and Sentinel-2 Satellite Data , 2020, Remote. Sens..

[56]  Laurence Shalloo,et al.  Development and adoption of new technologies to increase the efficiency and sustainability of pasture-based systems , 2015 .

[57]  E. Ruelle,et al.  Weather forecasts to enhance an Irish grass growth model , 2019, European Journal of Agronomy.

[58]  R. M. Sulc,et al.  Seasonal Variation in the Rising Plate Meter Calibration for Forage Mass , 2012 .

[59]  Mathew Legg,et al.  Ultrasonic Arrays for Remote Sensing of Pasture Biomass , 2019, Remote. Sens..

[60]  Silvia Valero,et al.  Challenges in Grassland Mowing Event Detection with Multimodal Sentinel Images , 2019, 2019 10th International Workshop on the Analysis of Multitemporal Remote Sensing Images (MultiTemp).

[61]  Jérôme Bindelle,et al.  Can Low-Cost Unmanned Aerial Systems Describe the Forage Quality Heterogeneity? Insight from a Timothy Pasture Case Study in Southern Belgium , 2020, Remote. Sens..

[62]  P. Carrère,et al.  Model predicting dynamics of biomass, structure and digestibility of herbage in managed permanent pastures. 1. Model description , 2006 .

[63]  Senén Barro,et al.  An extensive experimental survey of regression methods , 2019, Neural Networks.

[64]  Edzer Pebesma,et al.  Simple Features for R: Standardized Support for Spatial Vector Data , 2018, R J..

[65]  D. Nuyttens,et al.  Canopy height measurements and non‐destructive biomass estimation of Lolium perenne swards using UAV imagery , 2019, Grass and Forage Science.

[66]  Laurence Shalloo,et al.  PastureBase Ireland: A grassland decision support system and national database , 2017, Comput. Electron. Agric..

[67]  V. Vitale,et al.  Grassland Height Assessment by Satellite Images , 2017 .

[68]  Rachel Lugassi,et al.  Spatial and Temporal Monitoring of Pasture Ecological Quality: Sentinel-2-Based Estimation of Crude Protein and Neutral Detergent Fiber Contents , 2019, Remote. Sens..

[69]  Marie-Josée Cros.  SEPATOU: a Decision Support System for the Management of Rotational Grazing in a Dairy Production , 2001 .

[70]  M. Wachendorf,et al.  Combining ultrasonic sward height and spectral signatures to assess the biomass of legume-grass swards , 2013 .

[71]  Claudia Notarnicola,et al.  Review of Machine Learning Approaches for Biomass and Soil Moisture Retrievals from Remote Sensing Data , 2015, Remote. Sens..

[72]  C. Clark,et al.  Pasture growth model to assist management on dairy farms: Testing the concept with farmers , 2013 .

[73]  Fiona Cawkwell,et al.  Modeling Managed Grassland Biomass Estimation by Using Multitemporal Remote Sensing Data—A Machine Learning Approach , 2017, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[74]  Priyakant Sinha,et al.  Review of the use of remote sensing for biomass estimation to support renewable energy generation , 2015 .