Artificial Neural Network Model of Soil Heat Flux over Multiple Land Covers in South America

Andrade, Bruno César Comini de; Pedrollo, Olavo Correa; Ruhoff, Anderson Luís; Moreira, Adriana Aparecida; Laipelt, Leonardo; Kayser, Rafael Henrique Bloedow; Biudes, Marcelo Sacardi; Santos, Carlos Antonio Costa dos; Roberti, Débora Regina; Machado, Nadja Gomes; Dalmagro, Higo J.; Antonino, Antônio Celso Dantas; Lima, José Romualdo de Sousa; Souza, Eduardo Soares de; Souza, Rodolfo

doi:10.3390/rs13122337

Cited by 8 publications

(11 citation statements)

References 82 publications

(104 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…9,37-40 The G was the component of the energy balance estimated by SEBAL that had the lowest performance, given the weak correlation between the measured and estimated G. Despite this, their means did not differ statistically, and the error values were much smaller than the range of 30 to 80 Wm −2 obtained in forest, grassland, and urban areas. 9,[41][42][43][44] This is because G are related to surface coverage, and the plant canopy works as a thermal insulator, preventing the solar radiation reaches the ground. 15,41 Thus, G is dependent on the surface water availability and vegetation density, and therefore, point measurements of G should not be extrapolated to the entire area.…”

Section: Spatio-temporal Patterns Of Surface Variablesmentioning

confidence: 99%

Effect of urbanization on energy balance and evapotranspiration in an Amazon–Cerrado transition region in Brazil

Pinheiro

Biudes

Machado

et al. 2022

J. Appl. Rem. Sens.

Self Cite

View full text Add to dashboard Cite

.The Amazon–Cerrado transition region has a high rate of deforestation for agriculture, livestock, logging, and for the increase in urban areas of cities. The development of cities in this region in recent decades has increased the deforestation process and modified the local microclimate and the exchange of energy and water between the surface and the atmosphere. However, the magnitude of the change in energy and water exchanges due to urbanization is unknown for most Brazilian cities, especially in the Amazon–Cerrado region. Thus, the objective was to evaluate the spatiotemporal change in energy balance and in evapotranspiration (ET) patterns over three decades of urbanization in Sinop. The surface energy balance (SEB) and ET were estimated by the SEB algorithms for land with Landsat 5 Thematic Mapper images from the Sinop region, Mato Grosso, Brazil, in 1985, 2000, and 2010. The estimated SEB and ET were validated with measurements in a flux tower in the Amazon–Cerrado transition forest 50 km from the urban area of Sinop. The components of the SEB, except the soil/storage heat flux (G), had a strong correlation, and the ET had a moderate correlation with measurements in the flux tower. The surface albedo (asup), land surface temperature (LST), sensible heat flux (H), and G of the study area increased over the years, while the normalized difference vegetation index, latent heat flux (LE), and ET decreased over the years. Buildup and grassland areas had lower values of NDVI, Rn, LE, and ET and higher values of asup, LST, H, and G than forest areas. Therefore, the urbanization of Sinop has decreased available energy for ET and increased the energy to heat the air and soil.

show abstract

Section: Spatio-temporal Patterns Of Surface Variablesmentioning

confidence: 99%

Effect of urbanization on energy balance and evapotranspiration in an Amazon–Cerrado transition region in Brazil

Pinheiro

Biudes

Machado

et al. 2022

J. Appl. Rem. Sens.

Self Cite

View full text Add to dashboard Cite

show abstract

“…ML also shown optimal results for the estimation of environmental variables [52], H and LE retrievals [e.g., 53], other SEB components such as Rn, H and LE [e.g., 54,55] or ET [56,57]. For the estimation of G, [32] modelled the heat flux with ANN and RS data, and [33] compared the ANN and two empirical equations based on NDVI, Ts and α. In accordance with our findings, in both of the works, ML outperformed the other methods for the estimation of G. The underestimation of the empirical equations and the large errors found at high Fv ranges are partly because the majority of the type I and II equations rely on a single linear relationship with the predictor variables.…”

Section: Analysis Of the Accuracy And Performance Of The Different Me...mentioning

confidence: 99%

“…Nevertheless, up the date, only [32] modelled the G with RS data and the Artificial Neural Networks (ANN) ML algorithm. Furthermore, only [33] provided a comparison of the ANN model against two G/Rn empirical equations. To our best of knowledge, no study compared the performance of ensemble ML models, such as boosted regression trees (e.g., Random Forest (RF)), against the results obtained with neural networks (NN), and the ones obtained with the calibrated version of the G/Rn empirical equations.…”

Section: Introductionmentioning

confidence: 99%

Ensemble Machine Learning Outperforms Empirical Equations for the Ground Heat Flux Estimation with Remote Sensing Data

Bonsoms¹,

Boulet²

2022

Preprint

View full text Add to dashboard Cite

Estimating evapotranspiration at field scale is a major component of sustainable water management. Due to the difficulty to assess some major unknowns of the water cycle at that scale, including irrigation amounts, evapotranspiration is often computed as the residual of the instantaneous surface energy budget. One of the Surface Energy Bal-ance components with the largest uncertainties in their quantification over bare soils and sparse vegetation areas is the ground heat flux (G). Over the last decades, the es-timation of G with RS data has been mainly achieved with empirical equations, on the basis of the G and net radiation (Rn) ratio, G/Rn. G/Rn empirical equations generally require vegetation data (Type I empirical equations), in combination with surface tem-perature (Ts) and albedo (Type II empirical equations). In this article we aim to evalu-ate the estimation of G with RS. For the first time, we compare eight G/Rn empirical equations against two types of machine learning (ML) methods: an ensemble ML type, the Random Forest (RF), and the Neural Networks (NN). The comparison of each method is evaluated over dense dataset, including a wide range of climate and land covers, with data of Eddy-Covariance towers extended along the mid-latitude area that encompass the European and African continent. Our results have shown evidence that the driver of G in bare soils and sparse vegetation areas (Fraction of Vegetation, Fv <= 0.25) is Ts, instead of vegetation greenness indexes. On the other hand, the estimation of G with Rn, Ts or Fv decreases at dense vegetation areas (Fv >= 0.50). There are not significant differences between the most accurate type I and II empirical equations. For bare soils and sparse vegetation areas the empirical equation that better estimates G is E8, which combines the Leaf Area Index (LAI) and Ts. In dense vegetation areas (Fv >= 0.25), an exponential empirical equation based on Fv (E4), shows the best performance. However, ML better estimates G than the empirical equations, independently of the Fv ranges. A RF model with Rn, LAI and Ts as predictor variables shows the best accuracy and performance metrics, outperforming the NN model.

show abstract

“…Probably, the low performance of the G estimate is due to the low sensitivity of the model to the high spatial complexity of the study area. G tends not to have a high impact on the SEB and ET of densely vegetated surface, due to the lesser part of the available energy used to heat the soil, but G tends to impact the SEB and ET of surfaces with low vegetation cover, as the pastures and some natural grasslands in Cerrado and Pantanal [13,82,85].…”

Section: The Effects Of α and T S Retreival Models On Sebfs And Etmentioning

confidence: 99%

Surface Albedo and Temperature Models for Surface Energy Balance Fluxes and Evapotranspiration Using SEBAL and Landsat 8 over Cerrado-Pantanal, Brazil

Angelini

Biudes

Machado

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

The determination of the surface energy balance fluxes (SEBFs) and evapotranspiration (ET) is fundamental in environmental studies involving the effects of land use change on the water requirement of crops. SEBFs and ET have been estimated by remote sensing techniques, but with the operation of new sensors, some variables need to be parameterized to improve their accuracy. Thus, the objective of this study is to evaluate the performance of algorithms used to calculate surface albedo and surface temperature on the estimation of SEBFs and ET in the Cerrado-Pantanal transition region of Mato Grosso, Brazil. Surface reflectance images of the Operational Land Imager (OLI) and brightness temperature (Tb) of the Thermal Infrared Sensor (TIRS) of the Landsat 8, and surface reflectance images of the MODIS MOD09A1 product from 2013 to 2016 were combined to estimate SEBF and ET by the surface energy balance algorithm for land (SEBAL), which were validated with measurements from two flux towers. The surface temperature (Ts) was recovered by different models from the Tb and by parameters calculated in the atmospheric correction parameter calculator (ATMCORR). A model of surface albedo (asup) with surface reflectance OLI Landsat 8 developed in this study performed better than the conventional model (acon) SEBFs and ET in the Cerrado-Pantanal transition region estimated with asup combined with Ts and Tb performed better than estimates with acon. Among all the evaluated combinations, SEBAL performed better when combining asup with the model developed in this study and the surface temperature recovered by the Barsi model (Tsbarsi). This demonstrates the importance of an asup model based on surface reflectance and atmospheric surface temperature correction in estimating SEBFs and ET by SEBAL.

show abstract

Artificial Neural Network Model of Soil Heat Flux over Multiple Land Covers in South America

Cited by 8 publications

References 82 publications

Effect of urbanization on energy balance and evapotranspiration in an Amazon–Cerrado transition region in Brazil

Effect of urbanization on energy balance and evapotranspiration in an Amazon–Cerrado transition region in Brazil

Ensemble Machine Learning Outperforms Empirical Equations for the Ground Heat Flux Estimation with Remote Sensing Data

Surface Albedo and Temperature Models for Surface Energy Balance Fluxes and Evapotranspiration Using SEBAL and Landsat 8 over Cerrado-Pantanal, Brazil

Contact Info

Product

Resources

About