ABSTRACT
As per the Food and Agriculture Organization (FAO), agricultural productivity needs to be increased by 70% to feed a projected 10 billion people by the year 2050, and fertilizer application plays a key role in achieving this goal. Throughout the world, fertilizer usage has significantly increased to improve crop yields. Unfortunately, several studies have shown that over 65% of the fertilizer is being wasted leading to various environmental problems such as nitrogen runoff into lakes, rivers, and oceans. In addition, the traditional practice of uniform fertilizer application without regard to field and/or crop conditions can result in wasted fertilizer. To address these challenges, we propose a simple Q-learning-based simulation tool for studying the dynamic application of fertilizer. Q-learning is particularly well suited for solving this problem, as the framework allows sensing the environment to characterize the site and crop conditions and determine the amount of fertilizer to apply. We used remote sensing data as a proxy for crop health monitoring. We used reward shaping to determine the optimal amount of fertilizer. We compared our framework with other popular deep-learning approaches Deep Q Network (DQN), Double Deep Q Network (DDQN), and Dueling Network (Due_N). Experimental results show that our approach is computationally efficient while matching or performing better than other approaches.
- Takuya Akiba, Shotaro Sano, Toshihiko Yanase, Takeru Ohta, and Masanori Koyama. 2019. Optuna: A Next-generation Hyperparameter Optimization Framework. In Proceedings of the 25rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.Google ScholarDigital Library
- Milkha S Aulakh and Sukhdev S Malhi. 2005. Interactions of nitrogen with other nutrients and water: Effect on crop yield and quality, nutrient use efficiency, carbon sequestration, and environmental pollution. Advances in agronomy 86 (2005), 341--409.Google Scholar
- ATMA Choudhury and IR Kennedy. 2005. Nitrogen fertilizer losses from rice soils and control of environmental pollution problems. Communications in soil science and plant analysis 36, 11-12 (2005), 1625--1639.Google Scholar
- J Cihlar, L St Laurent, and JA Dyer. 1991. Relation between the normalized difference vegetation index and ecological variables. Remote sensing of Environment 35, 2-3 (1991), 279--298.Google Scholar
- Adrian A Correndo, Nicolas Tremblay, Jeffrey A Coulter, Dorivar Ruiz-Diaz, David Franzen, Emerson Nafziger, Vara Prasad, Luiz H Moro Rosso, Kurt Steinke, Juan Du, et al. 2021. Unraveling uncertainty drivers of the maize yield response to nitrogen: A Bayesian and machine learning approach. Agricultural and Forest Meteorology 311 (2021), 108668.Google ScholarCross Ref
- Melissa A Cregger, Nate G McDowell, Robert E Pangle, William T Pockman, and Aimée T Classen. 2014. The impact of precipitation change on nitrogen cycling in a semi-arid ecosystem. Functional ecology 28, 6 (2014), 1534--1544.Google Scholar
- Mark Denbaly and Harry Vroomen. 1993. Dynamic fertilizer nutrient demands for corn: A cointegrated and error-correcting system. American Journal of Agricultural Economics 75, 1 (1993), 203--209.Google ScholarCross Ref
- Maarten Elferink and Florian Schierhorn. 2016. Global demand for food is rising. Can we meet it. Harvard Business Review 7, 04 (2016), 2016.Google Scholar
- Udesh Gunarathna, Renata Borovica-Gajic, Shanika Karunasekera, and Egemen Tanin. 2022. Dynamic graph combinatorial optimization with multi-attention deep reinforcement learning. In Proceedings of the 30th International Conference on Advances in Geographic Information Systems. 1--12.Google ScholarDigital Library
- SM Haefele, MCS Wopereis, MK Ndiaye, and MJ Kropff. 2003. A framework to improve fertilizer recommendations for irrigated rice in West Africa. Agricultural systems 76, 1 (2003), 313--335.Google Scholar
- H Insam, CC Mitchell, and JF Dormaar. 1991. Relationship of soil microbial biomass and activity with fertilization practice and crop yield of three ultisols. Soil Biology and Biochemistry 23, 5 (1991), 459--464.Google ScholarCross Ref
- Jie Jiang, Peter M Atkinson, Jiayi Zhang, Ruhua Lu, Youyan Zhou, Qiang Cao, Yongchao Tian, Yan Zhu, Weixing Cao, and Xiaojun Liu. 2022. Combining fixedwing UAV multispectral imagery and machine learning to diagnose winter wheat nitrogen status at the farm scale. European Journal of Agronomy 138 (2022), 126537.Google ScholarCross Ref
- Roger Knowles. 1982. Denitrification. Microbiological reviews 46, 1 (1982), 43--70.Google Scholar
- Hao Liang, Songjuan Gao, and Kelin Hu. 2020. Global sensitivity and uncertainty analysis of the dynamic simulation of crop N uptake by using various N dilution curve approaches. European Journal of Agronomy 116 (2020), 126044.Google ScholarCross Ref
- Kunpeng Liu, Xiaolin Li, Cliff C Zou, Haibo Huang, and Yanjie Fu. 2020. Ambulance dispatch via deep reinforcement learning. In Proceedings of the 28th International Conference on Advances in Geographic Information Systems. 123--126.Google ScholarDigital Library
- Bao-Luo Ma and Dilip Kumar Biswas. 2015. Precision nitrogen management for sustainable corn production. In Sustainable agriculture reviews. Springer, 33--62.Google Scholar
- Ujjwal Mahajan and Bharat Raj Bundel. 2016. Drones for normalized difference vegetation index (NDVI), to estimate crop health for precision agriculture: A cheaper alternative for spatial satellite sensors. In Proceedings of the International Conference on Innovative Research in Agriculture, Food Science, Forestry, Horticulture, Aquaculture, Animal Sciences, Biodiversity, Ecological Sciences and Climate Change (AFHABEC-2016), Delhi, India, Vol. 22.Google Scholar
- German Mandrini, Sotirios V Archontoulis, Cameron M Pittelkow, Taro Mieno, and Nicolas F Martin. 2022. Simulated dataset of corn response to nitrogen over thousands of fields and multiple years in Illinois. Data in Brief 40 (2022), 107753.Google ScholarCross Ref
- German Mandrini, Cameron Mark Pittelkow, Sotirios Archontoulis, David Kanter, and Nicolas F Martin. 2022. Exploring Trade-Offs Between Profit, Yield, and the Environmental Footprint of Potential Nitrogen Fertilizer Regulations in the US Midwest. Frontiers in plant science 13 (2022), 852116--852116.Google Scholar
- German Mandrini, Cameron M Pittelkow, Sotirios V Archontoulis, Taro Mieno, and Nicolas F Martin. 2021. Understanding differences between static and dynamic nitrogen fertilizer tools using simulation modeling. Agricultural Systems 194 (2021), 103275.Google ScholarCross Ref
- Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Andrei A Rusu, Joel Veness, Marc G Bellemare, Alex Graves, Martin Riedmiller, Andreas K Fidjeland, Georg Ostrovski, et al. 2015. Human-level control through deep reinforcement learning. nature 518, 7540 (2015), 529--533.Google Scholar
- Emerson Nafziger. 2019. Managing Nitrogen for Corn in 2019. https://farmdocdaily.illinois.edu/wp-content/uploads/2019/03/fdd290319.pdfGoogle Scholar
- Rasel Parvej. 2021. Corn Nitrogen Management. https://www.lsuagcenter.com/articles/page1616180617871Google Scholar
- Nathalie Pettorelli, Jon Olav Vik, Atle Mysterud, Jean-Michel Gaillard, Compton J Tucker, and Nils Chr Stenseth. 2005. Using the satellite-derived NDVI to assess ecological responses to environmental change. Trends in ecology & evolution 20, 9 (2005), 503--510.Google Scholar
- Hanna J Poffenbarger, Daniel W Barker, Matthew J Helmers, Fernando E Miguez, Daniel C Olk, John E Sawyer, Johan Six, and Michael J Castellano. 2017. Maximum soil organic carbon storage in Midwest US cropping systems when crops are optimally nitrogen-fertilized. PLoS One 12, 3 (2017), e0172293.Google ScholarCross Ref
- Laila A Puntel, Agustin Pagani, and Sotirios V Archontoulis. 2019. Development of a nitrogen recommendation tool for corn considering static and dynamic variables. European Journal of Agronomy 105 (2019), 189--199.Google ScholarCross Ref
- Divina Gracia P Rodriguez, David S Bullock, and Maria A Boerngen. 2019. The origins, implications, and consequences of yield-based nitrogen fertilizer management. Agronomy Journal 111, 2 (2019), 725--735.Google ScholarCross Ref
- Yuji Saikai, Vivak Patel, and Paul D Mitchell. 2020. Machine learning for optimizing complex site-specific management. Computers and Electronics in Agriculture 174 (2020), 105381.Google ScholarCross Ref
- Sebastian Schmoll and Matthias Schubert. 2018. Vision paper: reinforcement learning in smart spatio-temporal environments. In Proceedings of the 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems. 81--84.Google ScholarDigital Library
- Shai Sela, Harold Mathijs van Es, Bianca N Moebius-Clune, Rebecca Marjerison, and Greg Kneubuhler. 2018. Dynamic model-based recommendations increase the precision and sustainability of N fertilization in midwestern US maize production. Computers and Electronics in Agriculture 153 (2018), 256--265.Google ScholarCross Ref
- Sadao Shoji, AT Gandeza, and K Kimura. 1991. Simulation of crop response to polyolefin-coated urea: II. Nitrogen uptake by corn. Soil Science Society of America Journal 55, 5 (1991), 1468--1473.Google ScholarCross Ref
- George Stanford. 1966. Nitrogen requirements of crops for maximum yield. Agricultural anhydrous ammonia technology and use (1966), 237--257.Google Scholar
- George Stanford. 1973. Rationale for optimum nitrogen fertilization in corn production. Journal of Environmental Quality 2, 2 (1973), 159--166.Google ScholarCross Ref
- PJ Thorburn, JS Biggs, SJ Attard, and Joseph Kemei. 2011. Environmental impacts of irrigated sugarcane production: nitrogen lost through runoff and leaching. Agriculture, ecosystems & environment 144, 1 (2011), 1--12.Google Scholar
- Sergio Adolfo Uhart and Fernando Héctor Andrade. 1995. Nitrogen defeciency in maize: I. Effects on crop growth, development, dry matter partitioning, and kernel set. Crop science 35, 5 (1995), 1376--1383.Google Scholar
- Hado Van Hasselt, Arthur Guez, and David Silver. 2016. Deep reinforcement learning with double q-learning. In Proceedings of the AAAI conference on artificial intelligence, Vol. 30.Google ScholarCross Ref
- Michael Vogt and Benjamin Rosman. 2022. Analyzing Reinforcement Learning Algorithms for Nitrogen Fertilizer Management in Simulated Crop Growth. In Proceedings of 43rd Conference of the South African Insti, Vol. 85. 146--157.Google Scholar
- Jordon Wade, Steve W Culman, Jessica AR Logan, Hanna Poffenbarger, M Scott Demyan, John H Grove, Antonio P Mallarino, Joshua M McGrath, Matthew Ruark, and Jaimie R West. 2020. Improved soil biological health increases corn grain yield in N fertilized systems across the Corn Belt. Scientific reports 10, 1 (2020), 3917.Google Scholar
- Min Wang, Qingsong Zheng, Qirong Shen, and Shiwei Guo. 2013. The critical role of potassium in plant stress response. International journal of molecular sciences 14, 4 (2013), 7370--7390.Google ScholarCross Ref
- Ruoyu Wang, Keith Cherkauer, and Laura Bowling. 2016. Corn response to climate stress detected with satellite-based NDVI time series. Remote Sensing 8, 4 (2016), 269.Google ScholarCross Ref
- Xinbing Wang, Yuxin Miao, Rui Dong, Hainie Zha, Tingting Xia, Zhichao Chen, Krzysztof Kusnierek, Guohua Mi, Hong Sun, and Minzan Li. 2021. Machine learning-based in-season nitrogen status diagnosis and side-dress nitrogen recommendation for corn. European Journal of Agronomy 123 (2021), 126193.Google ScholarCross Ref
- Christopher John Cornish Hellaby Watkins. 1989. Learning from delayed rewards. (1989).Google Scholar
- Jing Wu, Ran Tao, Pan Zhao, Nicolas F Martin, and Naira Hovakimyan. 2022. Optimizing Nitrogen Management with Deep Reinforcement Learning and Crop Simulations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 1712--1720.Google ScholarCross Ref
Index Terms
- Q-learning Based Simulation Tool for Studying Effectiveness of Dynamic Application of Fertilizer on Crop Productivity
Recommendations
Crop type mapping in a highly fragmented and heterogeneous agricultural landscape
This study provides a basis for better crop type mapping in central Iran.Based on phenological field-surveyed data, a set of Landsat 8 data is proposed.Novel decision tree classifiers along with SVM were used to classify crop types.Classification ...
A national knowledge-based crop recognition in mediterranean environment
Population growth, urban expansion, land degradation, civil strife and war may place plant natural resources for food and agriculture at risk. Crop and yield monitoring is basic information necessary for wise management of these resources. Satellite ...
Monitoring crop phenology with street-level imagery using computer vision
Highlights- Method to automatically collect side looking imagery for agriculture parcels.
- Crop types and phenology obtained from computer vision.
- Demonstration over monthly surveyed totalling 400,000 pictures for 17 crop types.
- Deep ...
AbstractStreet-level imagery holds a significant potential to scale-up in-situ data collection. This is enabled by combining the use of cheap high-quality cameras with recent advances in deep learning compute solutions to derive relevant thematic ...
Comments