ارزیابی روش‌های یادگیری ماشین در ریزمقیاس نمایی مکانی میانگین سالانة دمای سطح زمین و دمای هوا

نوع مقاله : مقاله پژوهشی

نویسندگان

1 علوم و مهندسی محیط زیست، دانشکده جغرافیا و علوم محیطی، دانشگاه حکیم سبزواری، سبزوار، ایران

2 گروه سنجش از دور و سیستم اطلاعات جغرافیایی، دانشکده جغرافیا و علوم محیطی، دانشگاه حکیم سبزواری، سبزوار، ایران

چکیده

امروزه استفاده از داده‌های شبکه‌ای پایگاه‌های اقلیمی مانند WorldClim یکی از منابع معتبر داده است که جایگزین داده‌های نقطه‌ای ایستگاه‌های هواشناسی شده است؛ اما استفاده از این پایگاه‌های اقلیمی باقدرت تفکیک مکانی پایین موجب ایجاد محدودیت برای بسیاری از مطالعات مرتبط با علوم زیست‌شناسی و بوم‌شناسی شده است. هدف از این پژوهش بررسی ارتباط دمای هوا و دمای سطح زمین و سپس بازتولید دمای سطح زمین باقدرت تفکیک مکانی بالا جهت ریزمقیاس نمایی میانگین سالانة دمای هوا با استفاده از دو محصول پرکاربرد میانگین سالانة دمای هوا از پایگاه داده WorldClim و میانگین سالانة دمای روز و شب سطح زمین MOD11A2 v061 سنجندة مادیس است. در این پژوهش، ابتدا عملکرد مدل‌های یادگیری ماشین شامل جنگل تصادفی،شبکة عصبی مصنوعی، رگرسیون شبکه الاستیک و ماشین بردار پشتیبان جهت ریزمقیاس نمایی محصول MOD11A2 v061 از 1 کیلومتر به 250 متر بررسی شد. برای این منظور از متغیرهای پیوسته و گسسته شامل ارتفاع از سطح دریا، عرض جغرافیایی، پوشش گیاهی، بافت خاک، جهت شیب و پوشش سطح زمین استفاده گردید. سپس میانگین سالانة دمای هوا WorldClim با استفاده از دمای سطح زمین با مدل پولی نومیال درجة 3  از 1 کیلومتر به 250 متر ریزمقیاس شد. همچنین از داده ‏های 7 ایستگاه سینوپتیک جهت بررسی اعتبار محصول ریزمقیاس شده استفاده شد. نتایج نمودار تیلور نشان داد مدل جنگل تصادفی، بهترین عملکرد در ریزمقیاس نمایی میانگین سالانة دمای سطح زمین با ریشة میانگین مربعات خطا 0/54 درجه سلسیوس دارد. همچنین مدل پولی نومیال درجة 3 میزان خطای نسبی کمتر در تولید داده ریزمقیاس دمای هوا دارد. مقدار ریشة میانگین مربعات خطا نتایج برای مدل تصحیح ‌نشده و تصحیح شده ریزمقیاس به ترتیب 1/32 و 1/21 درجه سلسیوس به‌دست آمد که با توجه به آزمون t جفتی اختلاف معنی‌داری در سطح 0/05 نشان نداد. یافته های این پژوهش نشان می‌دهد که ریزمقیاس نمایی میانگین سالانه دمای هوا WorldClim از اعتبار لازم برخوردار است.

کلیدواژه‌ها

عنوان مقاله [English]

Evaluation of machine learning methods in spatial downscaling of average annual land surface temperature and air temperature

نویسندگان [English]

  • Azadeh Atabati 1
  • Hamed Adab 2

1 Department of Environmental Sciences and Engineering, Faculty of Geography and Environmental Sciences, Hakim Sabzevari University, Sabzevar, Iran

2 Department of Remote Sensing and Geographic Information System, Faculty of Geography and Environmental Sciences, Hakim Sabzevari University, Sabzevar, Iran

چکیده [English]

Today, the use of raster data from climate databases such as WorldClim is one of the reliable data sources that have been used instead of the point data of weather stations. However, the use of these climate databases with low spatial resolution has generated limitations for many studies related to biological and ecological studies. This study aims to investigate the relationship between air temperature and land surface temperature and then reproduce the land surface temperature with the high spatial resolution for downscaling of annual average air temperature using two widely used products, namely,  annual average air temperature from the WorldClim database and annual average day and night temperature from MOD11A2 v061 MODIS sensor. In this study, firstly, the performance of machine learning models including random forest, artificial neural network, elastic network regression, and support vector machine for downscaling of MOD11A2 v061 product from 1 km to 250 meters was assessed. For this purpose, continuous and discrete predictor variables including height above sea level, latitude, vegetation cover, soil texture, slope direction, and land cover were used. Then, WorldClim's annual average air temperature was downscaled from 1 km to 250 meters using the land surface temperature with a 3rd-degree polynomial regression model.  Also, the weather data of seven synoptic stations were used to check the validity of the downscaled product. The results of the Taylor diagram represented that the random forest model has the best performance for the downscaled land surface temperatures with a root mean square error of 0.54 degrees Celsius. Also, the 3rd-degree polynomial regression model has a lower relative error rate in producing downscaled air temperature. The value of the root mean square error of the results for the uncorrected and corrected downscaled air temperature was 1.32 and 1.21 degrees Celsius, respectively, which did not show a significant difference at the 0.05 level according to the paired t-test. The findings of this research show that the downscaling of the mean annual air temperature of WorldClim has the required validity.

کلیدواژه‌ها [English]

  • WorldClim database
  • MODIS data
  • Spatial downscaling
  • Machine learning
Acharjee, A., Finkers, R., Visser, R.G., Maliepaard, C.J.M., 2013. Comparison of regularized regression methods for omics data. Metabolomics 3(3), 1-9.
Amiri, M., Tarkesh, M., Jafari, R., Jetschke, G., 2020. Bioclimatic variables from precipitation and temperature records vs. remote sensing-based bioclimatic variables: Which side can perform better in species distribution modeling?. Ecological Informatics 57, 1-45.
Anguita, D., Ghelardoni, L., Ghio, A., Oneto, L., Ridella. S., 2012. The ‘K’in K-fold cross validation. In: 20th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN) Bruges, Belgium, pp. 441-446.
Baldocchi, D., Ma, S., 2013. How will land use affect air temperature in the surface boundary layer? Lessons learned from a comparative study on the energy balance of an oak savanna and annual grassland in California, USA. Tellus B: Chemical and Physical Meteorology 65(1), 2-9.
Barry, R.G., 1992. Mountain weather and climate. Third Edition. Routledge, London, pp. 1-49.
Bartkowiak, P., Castelli, M., Notarnicola, C., 2019. Downscaling Land Surface Temperature from MODIS Dataset with Random Forest Approach over Alpine Vegetated Areas. Remote Sensing 11(11), 1-19.
Bollapragada, R., Nocedal, J., Mudigere D., Shi, H.-J., Tang, P.T.P., 2018. International Conference on Machine Learning (ICML). A progressive batching L-BFGS method for machine learning, Stockholm, Sweden. pp. 1-10.
Breiman, L., J. Friedman, H., Olshen, R.A., Stone, C.G. 1984. Classification and regression trees. Wadsworth International Group, Belmont, California, USA. 16(3), 199-215.
Chalghaf, B., Chemkhi, J., Mayala, B., Harrabi, M., Benie, G.B., Michael, E., Ben Salah, A., 2018. Ecological niche modeling predicting the potential distribution of Leishmania vectors in the Mediterranean basin: impact of climate change. Parasites  Vectors 11(461), 1-9.
Collados-Lara, A.-J., Fassnacht,S. R., Pardo-Igúzquiza, E.,  Pulido-Velazquez, D.., 2021. Assessment of high resolution air temperature fields at rocky mountain national park by combining scarce point measurements with elevation and remote sensing data. Remote Sensing 13(1), 1-26.
Daly, C., Taylor, G. H.,., 1998. The prism approach to mapping precipitation and temperature. 10th Conference on Applied Climatology and Workshop on Extreme Value Analysis in Climatology 1-4.
Didan, K., 2021. MODIS/Terra Vegetation Indices 16-Day L3 Global 250m SIN Grid V061. N. E. L. P. DAAC (Ed.). https://www.earthdata.nasa.gov/Accessed 20th February 2021
Duan, Q., Schaake, J.,  Koren, V., 2001. A priori estimation of land surface model parameters. in: land surface hydrology, meteorology, and climate: observations and modeling. American Geophysical Union 1-215.
Đurđević Babić, I., 2015. Predicting student satisfaction with courses based on log data from a virtual learning environment–a neural network and classification tree model. Croatian Operational Research Review 6(1), 105-120.
Ebrahimy, H., Aghighi, H., Azadbakht, M., Amani, M., Mahdavi, S.,  Matkan, A.A., 2021. Downscaling MODIS Land Surface Temperature Product Using an Adaptive Random Forest Regression Method and Google Earth Engine for a 19-Years Spatiotemporal Trend Analysis Over Iran. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14(1), 2103 - 2112.
Fick, S. E., Hijmans, R. J ., 2017. WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37(12), 4302-4315.
Ghorbanian, A., Kakooei,M., Amani, M., Mahdavi,S., Mohammadzadeh, A., Hasanlou, M., 2020. Improved land cover map of Iran using Sentinel imagery within Google Earth Engine and a novel automatic workflow for land cover classification using migrated training samples. ISPRS Journal of Photogrammetry and Remote Sensing 167, 276-288.
Grimm, R., Behrens, T., Märker, M.,  Elsenbeer, H., 2008. Soil organic carbon concentrations and stocks on Barro Colorado Island-Digital soil mapping using Random Forests analysis. Geoderma 146 (1-2), 102-113.
Hamann, A., Wang,  T., 2006. Potential effects of climate change on ecosystem and tree species distribution in British Columbia. Ecology 87, 2773-2786.
Han, F., Yan, J.,  Ling, H., 2021. Variance of vegetation coverage and its sensitivity to climatic factors in the Irtysh River basin. Peer J 9, 1-24.
He, J., Zhao,W., Li, A., Wen, F.,  Yu, D., 2019. The impact of the terrain effect on land surface temperature variation based on Landsat-8 observations in mountainous areas. International Journal of Remote Sensing 40(5-6), 1808-1827.
Hengl, T., Mendes de Jesus, J., Heuvelink,G. B. M., Ruiperez Gonzalez, M., Kilibarda, M., Blagotić, A., Shangguan, W., Wright, M.N., Geng, X., Bauer-Marschallinger, B., Guevara, M. A., Vargas, R., MacMillan, R.A., Batjes, N.H., Leenaars, J.G.B., Ribeiro, E., Wheeler, I., Mantel, S.,  Kempen, B., 2017. SoilGrids250m: Global gridded soil information based on machine learning. Plos One 12(2), 1-40.
Hijmans, R.J., Cameron, S.E., Parra, J.L., . Jones, P.G., Jarvis, A., 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25, 1965-1978.
Hutengs, C.,  Vohland, M., 2016. Downscaling land surface temperatures at regional scales with random forest regression. Remote Sensing of Environment 178, 127-141.
Karger, D. N., Conrad, O., Böhner, J., Kawohl, T., Kreft, H., Soria-Auza, R.W., Zimmermann, N.E., Linder,  H. P.,  Kessler, M., 2017. Climatologies at high resolution for the earth’s land surface areas. Scientific Data 4, 170122.
Kloog, I., Nordio, F., Coull, B.A.,  Schwartz, J., 2014. Predicting spatiotemporal mean air temperature using MODIS satellite surface temperature measurements across the Northeastern USA. Remote Sensing of Environment 150, 132-139.
Kunkel, K.E., 1989. Simple procedures for extrapolation of humidity variables in the mountainous western united states. Journal of Climate 2, 656-669.
La Sorte, F.A., Butchart, S.H., Jetz, W.,  Böhning-Gaese, K., 2014. Range-wide latitudinal and elevational temperature gradients for the world's terrestrial birds: implications under global climate change. PloS one 9(5), 1-13.
Liaw, A.,  Wiener, M.J.R., 2002. Classification and regression by random Forest. R News 2-3, 18-22.
Liston, G.E., Elder, K., 2006. A meteorological distribution system for high-resolution terrestrial modeling (MicroMet). Journal of Hydrometeorology 7, 217-234.
Lookingbill, T.R., Urban, D.L., 2003. Spatial estimation of air temperature differences for landscape-scale studies in montane environments. Agricultural and Forest Meteorology 114 (3-4), 141-151.
Maeda, E.E., 2014. Downscaling MODIS LST in the East African mountains using elevation gradient and land-cover information. International Journal of Remote Sensing 35(9), 3094-3108.
McCutchan, M.H., Fox, D.G., 1986. Effect of elevation and aspect on wind, temperature and humidity  Journal of Applied Meteorology and Climatology 25(12), 1996-2013.
Melchiorre, C., Castellanos Abella, E.A., van Westen, C.J.,  Matteucci, M., 2011. Evaluation of prediction capability, robustness, and sensitivity in non-linear landslide susceptibility models, Guantánamo, Cuba. Computers  Geosciences 37 (4), 410- 425.
Militino, A. F., Ugarte, M.D.,  Montesino, M., 2019. Filling missing data and smoothing altered data in satellite imagery with a spatial functional procedure. Stochastic Environmental Research and Risk Assessment 33, 1737-1750.
Miraboutalebi, S.M., Kazemi, P.,  Bahrami, P., 2016. Fatty Acid Methyl Ester (FAME) composition used for estimation of biodiesel cetane number employing random forest and artificial neural networks: A new approach. Fuel 166, 143-151.
Mohandes, M.A., Halawani, T.O., Rehman, S.,  Hussain, A. A., 2004. Support vector machines for wind speed prediction. Renewable Energy 29(6), 939-947.
Moreno, A.,  Hasenauer, H., 2016. Spatial downscaling of European climate data. International Journal of Climatology 36(3), 1444-1458.
Mwakapeje, E.R., Ndimuligo, S.A., Mosomtai, G., Ayebare, S., Nyakarahuka, L., Nonga, H.E., Mdegela, R.H.,  Skjerve, E., 2019. Ecological niche modeling as a tool for prediction of the potential geographic distribution of Bacillus anthracis spores in Tanzania. International Journal of Infectious Diseases 79  142-151.
Pal, S.,  Ziaul, S., 2017. Detection of land use and land cover change and land surface temperature in English Bazar urban centre. The Egyptian Journal of Remote Sensing and Space Science, 20 (1), 125-145.
Pineda, E.  Lobo, J. M., 2009. Assessing the accuracy of species distribution models to predict amphibian species richness patterns. Journal of Animal Ecology 78(1), 182-190.
Pratumchart, K., Suwannatrai, K., Sereewong, C., Thinkhamrop, K., Chaiyos, J., Boonmars, T.,  Suwannatrai, A. T., 2019. Ecological niche model based on maximum entropy for mapping distribution of bithynia siamensis goniomphalos, first intermediate host snail of opisthorchis viverrini in thailand. Acta Tropica 193, 183-191.
Rehfeldt, G. E., Crookston, N.L., Sáenz-Romero,  C., Campbell, E.M., 2012. North American vegetation model for land-use planning in a changing climate: a solution to large classification problems. Ecological applications : a publication of the Ecological Society of America 22(1), 119-141.
Ronao, C. A., Cho, S.-B., 2016. Human activity recognition with smartphone sensors using deep learning neural networks. Expert Systems with Applications 59, 235-244.
Sandholt, I., Rasmussen, K.,  Andersen, J., 2002. A simple interpretation of the surface temperature/vegetation index space for assessment of surface moisture status. Remote Sensing of Environment 79 (2-3), 213-224.
Soria-Auza, R. W., Kessler, M., Bach, K., Barajas-Barbosa, P. M., Lehnert, M., Herzog, S. K.,  Böhner, J., 2010. Impact of the quality of climate models for modelling species occurrences in countries with poor climatic documentation: a case study from Bolivia. Ecological Modelling 221(8), 1221-1229.
Sun, Z., Guo, H., Li, X., Lu, L.,  Du, X., 2011. Estimating urban impervious surfaces from Landsat-5 TM imagery using multilayer perceptron neural network and support vector machine. J. of Applied Remote Sensing 5(1), 1-18.
Taylor, K. E., 2001. Summarizing multiple aspects of model performance in a single diagram. Journal of Geophysical Research: Atmospheres 106 (D7), 7183-7192.
Tran, D.X., Pla, F.,Latorre-Carmona, P., Myint, S.W., Caetano, M.,  Kieu, H.V., 2017. Characterizing the relationship between land use land cover change and land surface temperature. ISPRS Journal of Photogrammetry and Remote Sensing 124, 119-132.
Uca, T., Ekhwan, J., Othman, M., Amal Rosmini, A.,  Ansari Saleh, A., 2018. Daily suspended sediment discharge prediction using multiple linear regression and artificial neural network. Journal of Physics: Conference Series 954, 1-20.
Vancutsem, C., Ceccato, P., Dinku, T.,  Connor, S.J., 2010. Evaluation of MODIS land surface temperature data to estimate air temperature in different ecosystems over Africa. Remote Sensing of Environment 114(2), 449-465.
Vapnik, V., 1998. Statistical learning theory., 1998. Wiley, New York. pp. 25-40.
Vernay, C., Blanc, P.,  Pitaval, S., 2013. Characterizing measurements campaigns for an innovative calibration approach of the global horizontal irradiation estimated by HelioClim-3. Renewable Energy 57, 339-347.
Wan, Z., Hook, S.,  Hulley, G., 2021. MODIS/Terra land surface temperature/emissivity 8-day l3 global 1km sin grid v061. N. E. L. P. DAAC (Ed.). https://www.earthdata.nasa.gov/Accessed 20th February 2021
Wang, D.-C., Zhang, G.-L.,Pan, X.-Z., Zhao, Y.-G.,Zhao, M.-S.,  Wang, G.-F., 2012. Mapping soil texture of a plain area using fuzzy-c-means clustering method based on land surface diurnal temperature difference. Pedosphere 22(3), 394-403.
Wang, K., Sun, J., Cheng, G.,  Jiang, H., 2011. Effect of altitude and latitude on surface air temperature across the Qinghai-Tibet Plateau. Journal of Mountain Science 8, 808-816.
Wang, T., Hamann, A., Spittlehouse, D.,  Carroll, C., 2016. Locally downscaled and spatially customizable climate data for historical and future periods for North America. PloS one 11 6), 1-17.
Williamson, S.N., Hik, D.S., Gamon, J.A.,  Kavanaugh, J.L.,  Flowers, G.E., 2014. Estimating temperature fields from modis land surface temperature and air temperature observations in a sub-arctic alpine environment. Remote Sensing 6 (2), 946-963.
Xia, J., Kumta, A.S., 2010. Feedforward Neural Network trained by BFGS algorithm for modeling plasma etching of silicon carbide. IEEE Transactions on Plasma Science 38(2), 142-148.
Zevenbergen, L.W., Thorne, C.R., 1987. Quantitative analysis of land surface topography. Earth surface processes and landforms 12(1), 47-56.
Zhao, W. Duan, S.-B., 2020. Reconstruction of daytime land surface temperatures under cloud-covered conditions using integrated MODIS/Terra land products and MSG geostationary satellite data. Remote Sensing of Environment 247, 1-15.
Zupan, B., Demsar, J., 2008. Open-Source tools for data mining. Clinics in Laboratory Medicine 28(1), 37-54.