Subtropikal Bir Yaz Çayırında Meteorolojik Parametreler-Toprak Sıcaklığı İlişkileri: Fiziksel Tabanlı ve Veriye Dayalı Modelleme

Year 2023, Volume: 54 Issue: 2, 48 - 56, 23.05.2023

Jannatul Ferdaous Progga Mohammad Nur Hossain Khan M.g. Mostofa Amın

https://doi.org/10.5152/AUAF.2023.23126

Cited By: 1

Abstract

Toprak sıcaklığı topraktaki fiziksel, kimyasal ve biyolojik süreçleri etkilediğinden, farklı derinliklerdeki toprak sıcaklığı dinamikleri bilgisi tarım endüstrisi için çok önemlidir. Bu çalışmada siltli tın tekstür sınıfına ait toprakların farklı derinliklerinde meteorolojik parametreler ile sıcaklık arasındaki ilişkiler, fiziksel tabanlı HYDRUS-1D modeli ve bir doğrusal regresyon modeli (LRM) kullanılarak değerlendirilmiştir. Çalışma alanında 5, 10, 20, 30 ve 50 cm derinliğindeki toprak katmanlarının sıcaklık değerleri ile meteoroloji istasyonundan alınan en düşük ve en yüksek hava sıcaklığı, basınç, çiğ oluşum noktası, yağış, güneşlenme süresi, rüzgar hızı verileri kullanılmıştır. Girdi kombinasyonları için korelasyon hassasiyeti araştırılmıştır. Çalışma sonucunda ortalama mutlak yüzde hatası (OMYH) ve R2’ye dayalı kantitatif değerlendirmelerin hem LRM hem de HYDRUS-1D modellerinden elde edilen tahminlerin tatmin edici olduğunu göstermiştir. LRM modelinde 5, 10 ve 20 cm derinlik katmanlarındaki R2 değerlerinin sırasıyla 0,96, 0,94 ve 0,88 olduğu, HYDRUS-1D modelinde ise 0,85, 0,86 ve 0,78 olduğu tespit edilmiştir. Benzer şekilde OMYH değerleri 5, 10 ve 20 cm derinlik kademelerinde LRM için %0,81, %0,87 ve %1,05 iken, HYDRUS-1D modeli için %3,44, %2,87 ve %3,73 olarak hesaplanmıştır. Genel olarak modellerin doğruluğu toprak derinliğinin artmasıyla azalmış ve 30cm’den daha derin katmanlarda her iki modelin de toprak sıcaklığını doğru bir şekilde tahmin edemediği belirlenmiştir. Toprak derinliğinin 50 cm olduğu katmanlarda R2 ve OMYH değerleri LRM modeli için 0,55 ve %1,25, HYDRUS-1D modeli için ise 0,51 ve %4,13 olmuştur. Çalışma sonucunda ayrıca LRM modelinin HYDRUS-1D modelinden daha iyi performans gösterdiği, beş bağımsız değişkenin (ortalama hava sıcaklığı, maksimum nem, yağış, rüzgar hızı ve buharlaşma) yaz mevsimindeki toprak sıcaklığını önemli ölçüde etkilediği, her iki yönteminde 0-20 cm’lik toprak derinliğinde toprak sıcaklığını tahmin etmek için tatmin edici bir şekilde kullanılabileceği belirlenmiştir.

Keywords

Buharlaşma, nem, HYDRUS-1D, lineer regresyon modeli, rüzgar hızı

Meteorological Parameters–Soil Temperature Relations in a Sub-Tropical Summer Grassland: Physically-Based and Data-Driven Modeling

Abstract

The knowledge of soil temperature dynamics at different depths is paramount for the agricultural industry because soil temperature impacts the physical, chemical, and biological processes in soil. A relationship between meteorological parameters and temperature at different depths in silt loam soil was assessed by using a physically based HYDRUS-1D model and a linear regression model. Soil temperature at 5, 10, 20, 30, and 50 cm soil layers, minimum and maximum air temperature, air pressure, relative humidity, dew point, rainfall, sunshine duration, wind speed, and evaporation data collected at a weather station were used. The correlation sensitivity for the input combinations was investigated. The quantitative evaluation based on mean absolute percentage error and R2 showed that the predictions of both linear regression model and HYDRUS-1D models were satisfactory. The R2 values at 5, 10, and 20 cm depths were 0.96, 0.94, and 0.88 for linear regression model, and 0.85, 0.86, and 0.78, for HYDRUS-1D model, respectively. Similarly, the mean absolute percentage error values for linear regression model were 0.81%, 0.87%, and 1.05%, whereas 3.44%, 2.87%, and 3.73% at 5, 10, and 20 cm depths for HYDRUS-1D model, respectively. Generally, the accuracy of the models diminished with increasing the soil depth. At >30 cm soil depth, both models failed to estimate soil temperature accurately. The R2 and mean absolute percentage error values at 50 cm depth for linear regression model were 0.55% and 1.25% and 0.51% and 4.13% for HYDRUS-1D, respectively. The linear regression model performed better than the HYDRUS-1D model. Five independent variables (mean air temperature, maximum humidity, rainfall, wind speed, and evaporation) were found to significantly affect the summer-time soil temperature. Either of the methods can be used satisfactorily to predict soil temperature at 0–20 cm soil depth.

Keywords

Evaporation, humidity, HYDRUS-1D, linear regression model, wind speed

References

• Ali, M. H., Amin, M. G. M., & Islam, A. K. M. R. (2005). Reference crop evapotranspiration (ETo) over Bangladesh and its implication in crop planning. Journal of Bangladesh Agricultural University, 3(1), 139–147. [CrossRef]
• Amin, M. G. M., Al Mİnhaj, A. A., Bhowmİk, B., Islam, D., & Islam, M. N. (2022). Nitrogen and phosphorus leaching and vegetative growth of maize as affected by animal manure application. Eurasian Journal of Soil Science, 11(1), 17–24. [CrossRef]
• Amin, M. G. M., Lima, L. A., Rahman, A., Liu, J., & Jahangir, M. M. R. (2023). Dairy manure application effects on water percolation, nutrient leaching and rice yield under alternate wetting and drying irrigation. International Journal of Plant Production, 17(1), 95–107. [CrossRef]
• Amin, M. G. M., Minhaj, A. A., Islam, D., Bhowmik, B., Hasan, M. M., & Islam, M. N. (2021). Mulch and no-till impacts on nitrogen and phosphorus leaching in maize field under sub-tropic monsoon climate. Environmental Challenges, 5, 5100346. [CrossRef]
• Bilgili, M. (2010). Prediction of soil temperature using regression and artificial neural network models. Meteorology and Atmospheric Physics, 110(1–2), 59–70. [CrossRef]
• Chung, S. O., & Horton, R. (1987). Soil heat and water flow with a partial surface mulch. Water Resources Research, 23(12), 2175–2186. [CrossRef]
• Citakoglu, H. (2017). Comparison of artificial intelligence techniques for prediction of soil temperatures in Turkey. Theoretical and Applied Climatology, 130(1–2), 545–556. [CrossRef] Craney, T. A., & Surles, J. G. (2002). Model-dependent variance inflation factor cutoff values. Quality Engineering, 14(3), 391–403. [CrossRef]
• De Marsily, G. (1986). Quantitative hydrogeology. Paris School of Mines. De Vries, D. A. (1963). Thermal properties of soils. In W. R. van Wijk (Ed.). Physics of plant environment (pp. 210–235). North-Holland Publishing Co.
• Delbari, M., Sharifazari, S., & Mohammadi, E. (2019). Modeling daily soil temperature over diverse climate conditions in Iran—A comparison of multiple linear regression and support vector regression techniques. Theoretical and Applied Climatology, 135(3–4), 991–1001. [CrossRef].
• George, R. K. (2001). Prediction of soil temperature by using artificial neural networks algorithms. Nonlinear Analysis: Theory, Methods and Applications, 47(3), 1737–1748. [CrossRef]
• Hossein, Z., & Ahmed, F. F. (2017). Estimating soil temperature from meteorological data using extreme learning machine, artificial neural network and multiple linear regression models. Iranian Journal of Soil and Water Research, 51(4), 895–906.
• Jebamalar, A. S., Raja, S., Thambi, A., & Sunitha, B. S. J. (2012). Prediction of annual and seasonal soil temperature variation using an artificial neural network. Indian Journal of Radio and Space Physics, 41(1), 48–57.
• Kalogirou, S. A., & Florides, G. A. (2004). Measurements of ground temperature at various depths. https://ktisis.cut.ac.cy/bitstream/20.500.14279/2417/3/C55-PRT020-SET3.pdf
• Kanzari, S., Mariem, S. B., Nouna, B. B., & Ilahy, R. (2018). Comparison of Hydrus-1D to thermal dispersion model for water-heat transport in a semi-arid region of Tunisia. Asian Journal of Applied Sciences, 11(2), 64–70. [CrossRef]
• Karnieli, A., Agam, N., Pinker, R. T., Anderson, M., Imhoff, M. L., Gutman, G. G., Panov, N., & Goldberg, A. ( 2010). Use of NDVI and land surface temperature for drought assessment: Merits and limitations. Journal of Climate, 23(3), 618–633. [CrossRef]
• Kim, S., & Singh, V. P. (2014). Modeling daily soil temperature using data driven models and spatial distribution. Theoretical and Applied Climatology, 118(3), 465–479. [CrossRef]
• Kisi, O., Tombul, M., & Kermani, M. Z. (2015). Modeling soil temperatures at different depths by using three different neural computing techniques. Theoretical and Applied Climatology, 121(1–2), 377–387. [CrossRef]
• Kleissl, J., Moreno, H., Hendrickx, J. M. H., & Simunek, J. (2007). HYDRUS simulations of soil surface temperatures. In Detection and Remediation Technologies for Mines and Minelike Targets, XII(6553), 213–224. [CrossRef]
• Kourat, T., Smadhi, D., Mouhouche, B., Gourari, N., Mostofa Amin, M. G. M., & Bryant, C. R. (2021). Assessment of future climate change impact on rainfed wheat yield in the semi-arid Eastern High Plain of Algeria using a crop model. Natural Hazards, 107(3), 2175–2203. [CrossRef]
• Melesse, A. M., & Hanley, R. S. (2005). Artificial neural network application for multi-ecosystem carbon flux simulation. Ecological Modelling, 189(3–4), 305–314. [CrossRef]
• Menon, S. P., Bharadwaj, R., Shetty, P., Sanu, P., & Nagendra, S. ( 2017). Prediction of temperature using linear regression. In International Conference on Electrical, Electronics, Communication, Computer, and Optimization Techniques (ICEECCOT) (pp. 1–6). [CrossRef].
• Miles, J. (2014). Tolerance and variance inflation factor. Wiley statsref: Statistics reference. [CrossRef]
• Saito, H., Šimůnek, J., & Mohanty, B. P. (2006). Numerical analysis of coupled water, vapor, and heat transport in the vadose zone. Vadose Zone Journal, 5(2), 784–800. [CrossRef]
• Sándor, R., & Fodor, N. (2012). Simulation of soil temperature dynamics with models using different concepts. TheScientificWorldJournal, 2012, 590287. [CrossRef] Shein, E., Mady, A., & Ili’n, L. (2019). Validation of HYDRUS-1D for Predicting of Soil Moisture Content with Hysteresis Effect. Biogeosystem Technique, 6(1), 59–64.
• Šimůnek, J., Van Genuchten, M. T., & Šejna, M. (2005). The HYDRUS-1D software package for simulating the one dimensional movement of water, heat, and multiple solutes in variably-saturated media. University of California-Riverside Research Reports, 3, 1–240.
• Tabari, H., Sabziparvar, A., & Ahmadi, A. (2010). Comparison of artificial neural network and multivariate linear regression methods for estimation of daily soil temperature in an arid region. Meteorology and Atmospheric Physics, 110(3), 135–142.
• Wu, W., Tang, X. P., Guo, N. J, Yang, C., Liu, H. B., & Shang, Y. F. (2013). Spatiotemporal modeling of monthly soil temperature using artificial neural networks. Theoretical and Applied Climatology, 113(3–4), 481–494.
• Yadav, B., Krishnan, P., Shafeeq, P. M., Parihar, C. M., & Aggarwal, P. (2020). Modelling soil thermal regime in wheat using HYDRUS-2D under diversified maize-wheat-mungbean cropping system. Catena, 194, 104765.
• Yılmaz, T., Özbek, A., Yılmaz, A., & Büyükalaca, O. (2009). Influence of upper layer properties on the ground temperature distribution. Journal of Thermal Science and Technology, 29(2), 43–51.