Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network

Opeyemi Samuel Sajo; Philip Gbenro Oguntunde; Johnson Toyin Fasinmirin; Akindele Akinnagbe; Ayorinde Akinlabi Olufayo; Samuel Ohikhena Agele

doi:doi:10.11648/j.ajnna.20210702.11

Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network

Opeyemi Samuel Sajo, Philip Gbenro Oguntunde, Johnson Toyin Fasinmirin, Akindele Akinnagbe, Ayorinde Akinlabi Olufayo, Samuel Ohikhena Agele

Published in American Journal of Neural Networks and Applications (Volume 7, Issue 2)

Received: 22 June 2021 Accepted: 10 July 2021 Published: 23 August 2021

Views: Downloads:

Download PDF

Share This Article

Twitter
Linked In
Facebook

Abstract

Direct measurement of crop water use is difficult and labour intensive. In some cases, the technicalities involved can only be exploited by well-trained researchers. Therefore, estimating this important crop parameter from readily available climatic data by way of modelling will ease the burden of direct measurement. The aim of the study is to parameterize models of canopy conductance of rain-fed cocoa tree, suitable for inclusion in physically-based model for predicting water use of cocoa trees. To do this, Sap flow density was monitored in three cocoa trees (Forestaro cultivar group) at the eight (8) year old cocoa plantation of the Federal University of Technology, Akure, Nigeria (7° 18' 15.9"N, 5° 07' 32.3"E), from 8^th March 2018 to 7^th March 2019, covering the two seasons of the region. Cocoa tree transpiration was determined from the measured sap flow and fitted into a physically based model (PM) to derive canopy conductance used for modelling. To choose the best model that predicts canopy conductance (the stomata control of water transport) in cocoa trees, Vector Autoregressive Models (VAR), a multivariate time series model, and Long Short-Term Memory (LSTM) network, an Artificial Intelligence (AI) model were employed. The prediction power of the VAR model was assessed and visualized using the vars R package, while the LSTM model, a Recurrent Neural Network (RNN) algorithm was implemented using Python programming within Google COLAB jupyter notebook. Before modelling, data were tested for stationarity using the Augmented Dickey-Fuller test. While two-thirds of the data were used to train the models, the remaining one-third of the data were used to test the trained model. As VAR models were evaluated using R-squared and Root Mean Squared Error (RMSE), LSTM was evaluated by comparing the train loss and test loss, and also RMSE. VAR (with Adjusted R-Squared=0.11) is found not to be suitable to model the complex relationship between canopy conductance and climatic variables. Further iteration to exclude insignificant climatic variables from the VAR model did not also improve the model. However, LSTM with RMSE of 0.026 and having the test loss not dropping below the training loss was observed to perform better in modelling the canopy conductance of Cocoa. The result of the research further revealed that temporal dynamics of transpiration is complex and difficult to be defined by traditional regression. LSTM with a prediction accuracy of 97.4% could therefore be used for the prediction of cocoa canopy conductance.

Published in	American Journal of Neural Networks and Applications (Volume 7, Issue 2)
DOI	10.11648/j.ajnna.20210702.11
Page(s)	23-29
Creative Commons	This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.
Copyright	Copyright © The Author(s), 2021. Published by Science Publishing Group

Keywords

Cocoa, Canopy Conductance, Sap Flow, Transpiration, Recurrent Neural Network

References

[1]	Caminha, H., Silva, T., Rocha, A., and Lima, S. (2017). Estimating reference evapotranspiration using data mining prediction models and feature selection. Proceedings of the 19th International Conference on Enterprise Information Systems (ICEIS 2017), 1: 272–279.
[2]	Braga, D., Coelho da Silva, T. L., Rocha, A., Coutinho, G., Magalhaes, R. P., Guerra, P. T., and de Macˆedo, J. A. F. (2018). Time Series Forecasting for Purposes of Irrigation Management Process. SBC 33rd Brazilian Symposium on Databases (SBBD), Rio de Janeiro, RJ, Brazil, 217-222. http://sisdagro.inmet.gov.br/sisdagro/app/monitoramento/bhc.
[3]	Gautam, R., and Sinha, A. K. (2016). Time series analysis of reference crop evapotranspiration for Bokaro District, Jharkhand, India. Journal of Water and Land Development, 30 (VII-IX): 51-56.
[4]	Smith, T. G. (2017). Pyramid. ARIMA estimators for Python. MIT, USA.
[5]	Box, G. E., Jenkins, G. M., Reinsel, G. C., and Ljung, G. M. (2015). Time series analysis: forecasting and control. John Wiley & Sons.
[6]	Song, H., & Li, G. (2008). Tourism demand modelling and forecasting: A review of recent research. Tourism Management, 29 (2): 203-220.
[7]	Wang, F., Yu, C., Xiong, L., and Chang, Y. (2019). How can agricultural water use efficiency be promoted in China? A spatial-temporal analysis, Resources, Conservation and Recycling, 145: 411-418.
[8]	Nasrullah, M., Rizwanullah, M., Yu, X., Jo, H., Sohail, T. M., and Liang, L. (2021). Autoregressive distributed lag (ARDL) approach to study the impact of climate change and other factors on rice production in South Korea. Journal of Water and Climate Change, jwc2021030. https://doi.org/10.2166/wcc.2021.030.
[9]	Ravazzani, G., Ghilardi, M., Mendlik, T., Gobiet, A., Corbari, C., and Mancini, M. (2014). Investigation of Climate Change Impact on Water Resources for an Alpine Basin in Northern Italy: Implications for Evapotranspiration Modeling Complexity. PLoS ONE 9 (10): e109053. https://doi.org/10.1371/journal.pone.0109053.
[10]	Trajkovic, S., Todorovic, B., and Stankovic, M. (2003). Forecasting of Reference Evapotranspiration by Artificial Neural Network. Journal of Irrigation and Drainage Engineering, 129 (6).
[11]	Fathian, F., Fakheri-Fard, A., Ouarda, T. B. M. J. et al. (2019). Multiple streamflow time series modeling using VAR–MGARCH approach. Stochastic Environmental Research and Risk Assessment, 33: 407–425. https://doi.org/10.1007/s00477-019-01651-9.
[12]	Saigal, S., and Mehrotra, D. (2012). Performance Comparison of Time Series Data Using Predictive Data Mining Techniques. Advances in Information Mining, 4 (1): 57-66.
[13]	Pfaff, B. (2008). VAR, SVAR and SVEC Models: Implementation Within R Package Vars. Journal of Statistical Software, 27 (4): 1-32. doi: DOI: 10.18637/jss.v027.i04.
[14]	Landeras, G., Ortiz-Barredo, A., and Lopez, J. J. (2009). Forecasting weekly Evapotranspiration with ARIMA and Artificial Neural Network Models. Technical Paper, American Society of Civil Engineers (ASCE) Library. https://doi.org/10.1061/(ASCE)IR.1943-4774.0000008.
[15]	Ertel, W. (2017). Introduction to Artificial Intelligence, Second Edition. Springer International Publishing AG. DIO 10.1007/978-3-319-58487-4.
[16]	Dou, X., and Yang, Y. (2018). Evapotranspiration estimation using four different machine learning approaches in different terrestrial ecosystems. Computer and Electronics in Agriculture. Elsevier, 148: 95-106.
[17]	Burger, C. J., Dohnal, M., Kathrada, M., and Law, R. (2001). A practitioner’s guide to time-series methods for tourism demand forecasting: a case study of Durban, South Africa. Tourism Management, 22: 403-409.
[18]	Bishop, C. M. (2005). Neural Networks for Pattern Recognition. Oxford: Oxford University Press.
[19]	Vapnik, V., and Chervonenkis, A. (1964). A note on one class of perceptron. Kernel Machines.
[20]	Shrestha, N. K., and Shukla, S. (2015). Support vector machine based modeling of evapotranspiration using hydro-climatic variables in a sub-tropical environment, Agricultural and Forest Meteorology, 200: 172-184. https://doi.org/10.1016/j.agrformet.2014.09.025.
[21]	Mohammadi, B,. and Mehdizadeh, S. (2020). Modeling daily reference evapotranspiration via a novel approach based on support vector regression coupled with whale optimization algorithm. Agricultural Water Management, Elsevier, 237. https://doi.org/10.1016/j.agwat.2020.106145.
[22]	Chen, H., Huang, J. J., and McBean, E. (2020). Partitioning of daily evapotranspiration using a modified shuttle worth-wallace model, random Forest and support vector regression, for a cabbage farmland. Agricultural Water Management, Elsevier, 228. https://doi.org/10.1016/j.agwat.2019.105923. (https://www.sciencedirect.com/science/article/pii/S0378377419313034).
[23]	Wen, X., Si, J., He, Z., Wu, J., Shao, H., and Yu, H. (2015). Support-Vector-Machine-Based Models for Modeling Daily Reference Evapotranspiration with Limited Climatic Data in Extreme Arid Regions. Water Resources Management, 29: 3195–3209. https://doi.org/10.1007/s11269-015-0990-2
[24]	Chen, S. M., and Tanuwijaya, K. (2011). Multivariate fuzzy forecasting based on fuzzy time series and automatic clustering techniques. Expert Systems with Applications, 38 (8): 10594–10605. Retrieved from http://dx.doi.org/10.1016/j.eswa.2011.02.098.
[25]	Roy, D. K. (2021). Long Short-Term Memory Networks to Predict One-Step Ahead Reference Evapotranspiration in a Subtropical Climatic Zone. Environmental Processes, 8 (2), 911-941.
[26]	Xiang, Z., Yan, J., and Demir, I. (2020). A rainfall-runoff model with LSTM-based sequence-to-sequence learning. Water resources research, 56 (1), e2019WR025326.
[27]	Datascienceplus (2021). Predicting Irish electricity consumption with an LSTM neural network. Advanced Modeling in R: Predicting Irish electricity consumption with an LSTM neural network. https://datascienceplus.com/predicting-irish-electricity-consumption-with-an-lstm-neural-network/. Retrieved on 7th July 2021.
[28]	Granier, A. (1987). Evaluation of transpiration in a Douglas-fir stand by means of sap flow measurements. Tree Physiol. 3, 309–320.
[29]	Monteith, J. L., and Unsworth, M. H. (1990). Principles of Environmental Physics. Arnold, New York.
[30]	Phillips, N., Nagchaudhuri, A., and Oren, R. G. (1997). Time constant for water transport in loblolly pine trees estimated from time series of evaporative demand and stem sap flow. Trees 11: 412–419.
[31]	Rana, G., Katerji, N., and Lorenzi, F. (2005). Measurement and modeling of evapotranspiration of irrigated citrus orchard under Mediterranean conditions. Agric. For. Meteorol., 128: 199–209.
[32]	Stewart, J. B. (1988). Modelling surface conductance of pine forest. Agric. For. Meteorol., 43: 19–35.
[33]	Oguntunde, P. G., Fasinminrin, J. T., and van de Giesen, N. (2011). Influence of Tree Age and Variety on Allometric Characteristics and Water Use of Mangifera indica L. Growing in Plantation. Journal of Botany, Hindawi Publishing Corporation. doi: 10.1155/2011/824201.
[34]	Lu, P., Yunusa, I. A. M., Walker, R. R., and Muller, J. (2003). Regulation of canopy conductance and transpiration and their modeling in irrigated grapevines. Functional Plant Biology 30: 689–698.
[35]	Oguntunde, P. G., and Alatise M. O. (2007). Environmental regulation and modelling of cassava canopy conductance under drying root-zone soil water. Meteorology Applications, Royal Meteorology Society, 14: 245-252.
[36]	Hyndman, R. J., and Athanasopoulos, G. (2016). Stationarity and differencing. OTexts. Retrieved from https://www.otexts.org/fpp/8/1.
[37]	Hyndman, R. J., and Athanasopoulos, G. (2016). Vector autoregressions. Forecasting: principles and practice. OTexts. Retrieved from https://www.otexts.org/fpp/9/2.
[38]	Leach, L. F., and Henson, R. K. (2007). The use and impact of adjusted R2 effects in published regression research. Multiple Linear Regression Viewpoints, 33 (1): 1-11.
[39]	Mahdi, E., Xiao, K. J., and McLeod, I. A. (2015). Esam Mahdi, Ken Jinkun Xiao and A. Ian McLeod. CRAN. Retrieved from https://cran.r-project.org/web/packages/portes/portes.pdf.
[40]	Song, H., and Witt, S. F. (2006). Forecasting international tourist flows to Macau. Tourism Management, 27: 214-224.
[41]	Oguntunde, P. G., and van de Giesen, N. (2005). Water Flux measurement and prediction in young cashew trees using sap flow data. Hydrol. Processes 19: 3235–3248.

Cite This Article

Plain Text BibTeX RIS

APA Style

Opeyemi Samuel Sajo, Philip Gbenro Oguntunde, Johnson Toyin Fasinmirin, Akindele Akinnagbe, Ayorinde Akinlabi Olufayo, et al. (2021). Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network. American Journal of Neural Networks and Applications, 7(2), 23-29. https://doi.org/10.11648/j.ajnna.20210702.11

Copy | Download

ACS Style

Opeyemi Samuel Sajo; Philip Gbenro Oguntunde; Johnson Toyin Fasinmirin; Akindele Akinnagbe; Ayorinde Akinlabi Olufayo, et al. Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network. Am. J. Neural Netw. Appl. 2021, 7(2), 23-29. doi: 10.11648/j.ajnna.20210702.11

Copy | Download

AMA Style

Opeyemi Samuel Sajo, Philip Gbenro Oguntunde, Johnson Toyin Fasinmirin, Akindele Akinnagbe, Ayorinde Akinlabi Olufayo, et al. Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network. Am J Neural Netw Appl. 2021;7(2):23-29. doi: 10.11648/j.ajnna.20210702.11

Copy | Download

@article{10.11648/j.ajnna.20210702.11,
  author = {Opeyemi Samuel Sajo and Philip Gbenro Oguntunde and Johnson Toyin Fasinmirin and Akindele Akinnagbe and Ayorinde Akinlabi Olufayo and Samuel Ohikhena Agele},
  title = {Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network},
  journal = {American Journal of Neural Networks and Applications},
  volume = {7},
  number = {2},
  pages = {23-29},
  doi = {10.11648/j.ajnna.20210702.11},
  url = {https://doi.org/10.11648/j.ajnna.20210702.11},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ajnna.20210702.11},
  abstract = {Direct measurement of crop water use is difficult and labour intensive. In some cases, the technicalities involved can only be exploited by well-trained researchers. Therefore, estimating this important crop parameter from readily available climatic data by way of modelling will ease the burden of direct measurement. The aim of the study is to parameterize models of canopy conductance of rain-fed cocoa tree, suitable for inclusion in physically-based model for predicting water use of cocoa trees. To do this, Sap flow density was monitored in three cocoa trees (Forestaro cultivar group) at the eight (8) year old cocoa plantation of the Federal University of Technology, Akure, Nigeria (7° 18' 15.9"N, 5° 07' 32.3"E), from 8th March 2018 to 7th March 2019, covering the two seasons of the region. Cocoa tree transpiration was determined from the measured sap flow and fitted into a physically based model (PM) to derive canopy conductance used for modelling. To choose the best model that predicts canopy conductance (the stomata control of water transport) in cocoa trees, Vector Autoregressive Models (VAR), a multivariate time series model, and Long Short-Term Memory (LSTM) network, an Artificial Intelligence (AI) model were employed. The prediction power of the VAR model was assessed and visualized using the vars R package, while the LSTM model, a Recurrent Neural Network (RNN) algorithm was implemented using Python programming within Google COLAB jupyter notebook. Before modelling, data were tested for stationarity using the Augmented Dickey-Fuller test. While two-thirds of the data were used to train the models, the remaining one-third of the data were used to test the trained model. As VAR models were evaluated using R-squared and Root Mean Squared Error (RMSE), LSTM was evaluated by comparing the train loss and test loss, and also RMSE. VAR (with Adjusted R-Squared=0.11) is found not to be suitable to model the complex relationship between canopy conductance and climatic variables. Further iteration to exclude insignificant climatic variables from the VAR model did not also improve the model. However, LSTM with RMSE of 0.026 and having the test loss not dropping below the training loss was observed to perform better in modelling the canopy conductance of Cocoa. The result of the research further revealed that temporal dynamics of transpiration is complex and difficult to be defined by traditional regression. LSTM with a prediction accuracy of 97.4% could therefore be used for the prediction of cocoa canopy conductance.},
 year = {2021}
}

Copy | Download

TY  - JOUR
T1  - Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network
AU  - Opeyemi Samuel Sajo
AU  - Philip Gbenro Oguntunde
AU  - Johnson Toyin Fasinmirin
AU  - Akindele Akinnagbe
AU  - Ayorinde Akinlabi Olufayo
AU  - Samuel Ohikhena Agele
Y1  - 2021/08/23
PY  - 2021
N1  - https://doi.org/10.11648/j.ajnna.20210702.11
DO  - 10.11648/j.ajnna.20210702.11
T2  - American Journal of Neural Networks and Applications
JF  - American Journal of Neural Networks and Applications
JO  - American Journal of Neural Networks and Applications
SP  - 23
EP  - 29
PB  - Science Publishing Group
SN  - 2469-7419
UR  - https://doi.org/10.11648/j.ajnna.20210702.11
AB  - Direct measurement of crop water use is difficult and labour intensive. In some cases, the technicalities involved can only be exploited by well-trained researchers. Therefore, estimating this important crop parameter from readily available climatic data by way of modelling will ease the burden of direct measurement. The aim of the study is to parameterize models of canopy conductance of rain-fed cocoa tree, suitable for inclusion in physically-based model for predicting water use of cocoa trees. To do this, Sap flow density was monitored in three cocoa trees (Forestaro cultivar group) at the eight (8) year old cocoa plantation of the Federal University of Technology, Akure, Nigeria (7° 18' 15.9"N, 5° 07' 32.3"E), from 8th March 2018 to 7th March 2019, covering the two seasons of the region. Cocoa tree transpiration was determined from the measured sap flow and fitted into a physically based model (PM) to derive canopy conductance used for modelling. To choose the best model that predicts canopy conductance (the stomata control of water transport) in cocoa trees, Vector Autoregressive Models (VAR), a multivariate time series model, and Long Short-Term Memory (LSTM) network, an Artificial Intelligence (AI) model were employed. The prediction power of the VAR model was assessed and visualized using the vars R package, while the LSTM model, a Recurrent Neural Network (RNN) algorithm was implemented using Python programming within Google COLAB jupyter notebook. Before modelling, data were tested for stationarity using the Augmented Dickey-Fuller test. While two-thirds of the data were used to train the models, the remaining one-third of the data were used to test the trained model. As VAR models were evaluated using R-squared and Root Mean Squared Error (RMSE), LSTM was evaluated by comparing the train loss and test loss, and also RMSE. VAR (with Adjusted R-Squared=0.11) is found not to be suitable to model the complex relationship between canopy conductance and climatic variables. Further iteration to exclude insignificant climatic variables from the VAR model did not also improve the model. However, LSTM with RMSE of 0.026 and having the test loss not dropping below the training loss was observed to perform better in modelling the canopy conductance of Cocoa. The result of the research further revealed that temporal dynamics of transpiration is complex and difficult to be defined by traditional regression. LSTM with a prediction accuracy of 97.4% could therefore be used for the prediction of cocoa canopy conductance.
VL  - 7
IS  - 2
ER  -

Copy | Download

Author Information

Opeyemi Samuel Sajo

Department of Agricultural & Environmental Engineering, Federal University of Technology, Akure, Nigeria
Philip Gbenro Oguntunde

Department of Agricultural & Environmental Engineering, Federal University of Technology, Akure, Nigeria
Johnson Toyin Fasinmirin

Department of Agricultural & Environmental Engineering, Federal University of Technology, Akure, Nigeria
Akindele Akinnagbe

Division of Data Science, Te Whanau O Waipareira, Henderson, Auckland, New Zealand
Ayorinde Akinlabi Olufayo

Department of Agricultural & Environmental Engineering, Federal University of Technology, Akure, Nigeria
Samuel Ohikhena Agele

Department of Crop, Soil and Pest Management, Federal University of Technology, Akure, Nigeria

Download PDF

Sections

Plain Text BibTeX RIS

APA Style

Opeyemi Samuel Sajo, Philip Gbenro Oguntunde, Johnson Toyin Fasinmirin, Akindele Akinnagbe, Ayorinde Akinlabi Olufayo, et al. (2021). Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network. American Journal of Neural Networks and Applications, 7(2), 23-29. https://doi.org/10.11648/j.ajnna.20210702.11

Copy | Download

ACS Style

Opeyemi Samuel Sajo; Philip Gbenro Oguntunde; Johnson Toyin Fasinmirin; Akindele Akinnagbe; Ayorinde Akinlabi Olufayo, et al. Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network. Am. J. Neural Netw. Appl. 2021, 7(2), 23-29. doi: 10.11648/j.ajnna.20210702.11

Copy | Download

AMA Style

Opeyemi Samuel Sajo, Philip Gbenro Oguntunde, Johnson Toyin Fasinmirin, Akindele Akinnagbe, Ayorinde Akinlabi Olufayo, et al. Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network. Am J Neural Netw Appl. 2021;7(2):23-29. doi: 10.11648/j.ajnna.20210702.11

Copy | Download

@article{10.11648/j.ajnna.20210702.11,
  author = {Opeyemi Samuel Sajo and Philip Gbenro Oguntunde and Johnson Toyin Fasinmirin and Akindele Akinnagbe and Ayorinde Akinlabi Olufayo and Samuel Ohikhena Agele},
  title = {Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network},
  journal = {American Journal of Neural Networks and Applications},
  volume = {7},
  number = {2},
  pages = {23-29},
  doi = {10.11648/j.ajnna.20210702.11},
  url = {https://doi.org/10.11648/j.ajnna.20210702.11},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ajnna.20210702.11},
  abstract = {Direct measurement of crop water use is difficult and labour intensive. In some cases, the technicalities involved can only be exploited by well-trained researchers. Therefore, estimating this important crop parameter from readily available climatic data by way of modelling will ease the burden of direct measurement. The aim of the study is to parameterize models of canopy conductance of rain-fed cocoa tree, suitable for inclusion in physically-based model for predicting water use of cocoa trees. To do this, Sap flow density was monitored in three cocoa trees (Forestaro cultivar group) at the eight (8) year old cocoa plantation of the Federal University of Technology, Akure, Nigeria (7° 18' 15.9"N, 5° 07' 32.3"E), from 8th March 2018 to 7th March 2019, covering the two seasons of the region. Cocoa tree transpiration was determined from the measured sap flow and fitted into a physically based model (PM) to derive canopy conductance used for modelling. To choose the best model that predicts canopy conductance (the stomata control of water transport) in cocoa trees, Vector Autoregressive Models (VAR), a multivariate time series model, and Long Short-Term Memory (LSTM) network, an Artificial Intelligence (AI) model were employed. The prediction power of the VAR model was assessed and visualized using the vars R package, while the LSTM model, a Recurrent Neural Network (RNN) algorithm was implemented using Python programming within Google COLAB jupyter notebook. Before modelling, data were tested for stationarity using the Augmented Dickey-Fuller test. While two-thirds of the data were used to train the models, the remaining one-third of the data were used to test the trained model. As VAR models were evaluated using R-squared and Root Mean Squared Error (RMSE), LSTM was evaluated by comparing the train loss and test loss, and also RMSE. VAR (with Adjusted R-Squared=0.11) is found not to be suitable to model the complex relationship between canopy conductance and climatic variables. Further iteration to exclude insignificant climatic variables from the VAR model did not also improve the model. However, LSTM with RMSE of 0.026 and having the test loss not dropping below the training loss was observed to perform better in modelling the canopy conductance of Cocoa. The result of the research further revealed that temporal dynamics of transpiration is complex and difficult to be defined by traditional regression. LSTM with a prediction accuracy of 97.4% could therefore be used for the prediction of cocoa canopy conductance.},
 year = {2021}
}

Copy | Download

TY  - JOUR
T1  - Modelling the Canopy Conductance of Cocoa Tree Using a Recurrent Neural Network
AU  - Opeyemi Samuel Sajo
AU  - Philip Gbenro Oguntunde
AU  - Johnson Toyin Fasinmirin
AU  - Akindele Akinnagbe
AU  - Ayorinde Akinlabi Olufayo
AU  - Samuel Ohikhena Agele
Y1  - 2021/08/23
PY  - 2021
N1  - https://doi.org/10.11648/j.ajnna.20210702.11
DO  - 10.11648/j.ajnna.20210702.11
T2  - American Journal of Neural Networks and Applications
JF  - American Journal of Neural Networks and Applications
JO  - American Journal of Neural Networks and Applications
SP  - 23
EP  - 29
PB  - Science Publishing Group
SN  - 2469-7419
UR  - https://doi.org/10.11648/j.ajnna.20210702.11
AB  - Direct measurement of crop water use is difficult and labour intensive. In some cases, the technicalities involved can only be exploited by well-trained researchers. Therefore, estimating this important crop parameter from readily available climatic data by way of modelling will ease the burden of direct measurement. The aim of the study is to parameterize models of canopy conductance of rain-fed cocoa tree, suitable for inclusion in physically-based model for predicting water use of cocoa trees. To do this, Sap flow density was monitored in three cocoa trees (Forestaro cultivar group) at the eight (8) year old cocoa plantation of the Federal University of Technology, Akure, Nigeria (7° 18' 15.9"N, 5° 07' 32.3"E), from 8th March 2018 to 7th March 2019, covering the two seasons of the region. Cocoa tree transpiration was determined from the measured sap flow and fitted into a physically based model (PM) to derive canopy conductance used for modelling. To choose the best model that predicts canopy conductance (the stomata control of water transport) in cocoa trees, Vector Autoregressive Models (VAR), a multivariate time series model, and Long Short-Term Memory (LSTM) network, an Artificial Intelligence (AI) model were employed. The prediction power of the VAR model was assessed and visualized using the vars R package, while the LSTM model, a Recurrent Neural Network (RNN) algorithm was implemented using Python programming within Google COLAB jupyter notebook. Before modelling, data were tested for stationarity using the Augmented Dickey-Fuller test. While two-thirds of the data were used to train the models, the remaining one-third of the data were used to test the trained model. As VAR models were evaluated using R-squared and Root Mean Squared Error (RMSE), LSTM was evaluated by comparing the train loss and test loss, and also RMSE. VAR (with Adjusted R-Squared=0.11) is found not to be suitable to model the complex relationship between canopy conductance and climatic variables. Further iteration to exclude insignificant climatic variables from the VAR model did not also improve the model. However, LSTM with RMSE of 0.026 and having the test loss not dropping below the training loss was observed to perform better in modelling the canopy conductance of Cocoa. The result of the research further revealed that temporal dynamics of transpiration is complex and difficult to be defined by traditional regression. LSTM with a prediction accuracy of 97.4% could therefore be used for the prediction of cocoa canopy conductance.
VL  - 7
IS  - 2
ER  -

Copy | Download