Remote sensing based evapotranspiration modeling for sugarcane in Brazil using a hybrid approach
Introduction
Brazil expects to harvest about 10 million hectares of sugarcane for the 2020/2021 season, which corresponds to 665 million metric tons, establishing the world's largest producer of sugarcane. The State of São Paulo is the main contributor with an estimated production of more than 360 million tons total. The main products derived from sugarcane are biofuel and sugar, with 30 billion liters of ethanol and almost 42 million tons of sugar estimated for the next harvest. These numbers could be even higher, given that most of the cultivated areas are under rainfed conditions which can negatively impact sugarcane productivity in Brazil. Productivity is much lower than that achieved in irrigated areas, approximately 69 t ha−1, due to water stress that can occur in critical stages of crop development (CONAB, 2021). Due to the poor distribution of rain over the long sugarcane growing season (one year each season), the crop is exposed to rainy months and very dry months during its development. As a result, the expansion of the irrigated sugarcane area in Brazil has increased rapidly due to gains in stalk yield (ANA, 2020). Many studies carried out in Brazil have demonstrated that irrigation increases the productivity of sugarcane (Gonçalves et al., 2019, Dias and Sentelhas, 2019, Uribe et al., 2013).
The optimization of water use in irrigated agriculture has an important role in the economy, production of food, and environmental security to guarantee the profitability of agricultural activity, and the sustainability of water resources. This is especially important in regions that face water scarcity, as is the case of sugarcane areas in the State of São Paulo. Considering that the use of irrigation is becoming more prevalent for sugarcane production, more efficient and low-cost methods to estimate crop evapotranspiration are important to ensure the proper estimation of the irrigation depth, including models that provide suitable monitoring of water in the root system and crop development in sugarcane irrigated areas.
Evapotranspiration (ET) estimation, based on remote sensing methods using satellite data, has been used widely due to low cost/low impact techniques, high temporal, and spatial resolution sensors that can be used at field scales, as well as global scales, in agricultural systems both in Brazil and worldwide. Recent works such as Venancio et al. (2019), Foster et al. (2019), Gonçalves et al. (2020), and Campos et al. (2018) demonstrated how remote sensing (RS) can be used for the monitoring and management of water resources in irrigated agriculture.
Some RS-based methodologies for estimating ET are based on the energy balance approach using radiometric land surface temperature and meteorological variables to estimate ET as a residual surface energy balance component (SEB) and, these models can be one source when ET is estimated for soil and vegetation together as a single layer such as the well-known SEBAL (Bastiaanssen et al., 1998) and METRIC (Allen and Wright, 1997), and two sources when ET is estimated for soil and vegetation individually such as the Two Source Energy Balance (TSEB) by Norman et al. (1995), and ALEX-DisALEX model (Anderson et al., 1997, Mecikalski et al., 1999, Norman et al., 2003). These models use orbital imagery, and ground-based meteorological data and, depending on the model, also can use field data to estimate the components of the energy balance (net radiation, sensitive heat, and soil heat flux) and as a result, obtain latent heat flux or ET.
There are other methods to estimate ET based on the basal crop coefficient as a function of the spectral reflectance values (Kcbrf) derived from vegetation indexes (VI) such as NDVI (Normalized difference Vegetation Index) and SAVI (Soil Adjusted Vegetation Index). The Kcbrf values are used in determining the actual crop ET used to estimate the remote sensing-based soil water balance (RSWB), as described in FAO 56 (Allen et al., 1998). The Kcbrf approach (Neale et al., 1989) has been widely applied to several crops (Campos et al., 2016, Campos et al., 2017, González-Dugo et al., 2013, Neale et al., 2021, Gonçalves et al., 2020, Jayanthi et al., 2007). Also, new biophysical photosynthesis model based on RS such as STIC-RCEEP, integrating land surface temperature (TR)-based ET or latent heat flux (LE) into a Remote sensing-driven approach to coupling Ecosystem Evapotranspiration and Photosynthesis (RCEEP) model, advantages of STIC-RCEEP are prominent under dry conditions (Bai et al., 2022).
In this research, the hybrid model known as Spatial EvapoTranspiration Modeling Interface (SETMI) (Geli and Neale, 2012, Neale et al., 2012) was applied. The modeling approach is based on coupling the TSEB and Kcbrf approaches. The TSEB model provides estimates of actual crop ET while the Kcbrf (from SAVI corresponding to the time of the satellite overpass) approach allows for updating the basal crop coefficient as well as the interpolation and extrapolation of ET between satellite image acquisition dates, improving the maintenance of a soil water balance in the crop root zone. Additionally, SETMI considers three layers in the soil profile to estimate the RSWB considering the soil heterogeneity pixel by pixel, still, it allows to upload of the climatic input data as tables or raster grid format. Also, the variables such as the variables ET extrapolation, initial canopy temperature, wind adjustment methods, green fraction, canopy height, effective precipitation, basal crop coefficient progression and interpolation can be estimated using more than one method to meet different user needs to run the TSEB and RSWB in the SETMI. Also, SETMI has the ability to provide prescription irrigation maps from ET estimated based on the Kcbrf- water balance allowing temporal interpolation and extrapolation of a spatial water balance between input image dates, this approach has the potential to be used for real-time irrigation scheduling as described in Barker et al. (2018).
The Kcb-VI relationship for sugarcane under Brazilian tropical climate has not been developed yet, therefore, for the proper estimation of ET and RSWB for sugarcane using the SETMI hybrid model, it is necessary to develop a specific Kcb-VI relationship for sugarcane in such conditions. This research proposes to develop a Kcb-VI relationship through Kcb measured in the field with an eddy covariance flux tower and the SAVI vegetation index from satellite imagery for sugarcane grown in northwest state of Sao Paulo, Brazil. Additionally, after adjusting the model, the objective of this research also was to evaluate the performance results of the estimated energy balance (EB) components and ET using SETMI against field data obtained from the turbulent flux data (eddy covariance). After establishing the Kcb-VI relationship and validating TSEB from SETMI, the hybrid model was applied to estimate daily Kcb, Kc and water balance in the root zone focusing on monitoring the crop growth and irrigation management over two sugarcane seasons in Brazil.
Section snippets
Study site
The research was carried out in a commercial field of 24 ha, close to Andradina, State of São Paulo, Latitude 20°43'43.6″ S, Longitude 51º16'30.3″ W, 360 m of altitude (Fig. 1), grown with sugarcane for two ratoon seasons, fourth and fifth harvesting seasons respectively (June, 2016 to June, 2018) with the variety RB96–6928.
According to the Koppen classification, the climate of the region is defined as tropical type (Aw) with dry winter and rainy summer (Peel et al., 2007), with average annual
Results and discussion
Fig. 3 presents the daily precipitation and average daily temperature over the growing seasons 2016–2018 (fourth ratoon) and 2017–2018 (fifth ratoon) recorded for the flux tower set up in the field.
Overall, the fourth and fifth growing seasons presented the same average temperature equal to 24.6 °C, greater than the historical average for the region (23.3 °C). Accumulated precipitation was 988 mm and 974 mm in the fourth and fifth ratoon years respectively, lower than the average historical
Conclusions
This research was carried out in the largest sugarcane producing region of Brazil using the SETMI hybrid model to estimate evapotranspiration and rootzone water balance of a growing sugarcane field, over two growing seasons. The modeled EB components using the SETMI hybrid model agreed well with the eddy covariance system values. The Kcb-SAVI relationship for sugarcane presented a strong correlation with an R2 value of 0.85 and an "ρ" of 0.92. As result, the SETMI hybrid model can be applied to
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors thank to the UNESP Irrigation and Drainage Program and the Daugherty Water for Food Global Institute of the University of Nebraska for the scientific technical support and partial funding. Thanks also to CAPES (Process 88881.189165/2018-01), CNPq (Process 404.229/2013-1), and FAPESP (Process 2.009/52.467-4; 2020/08365-1) for the financial support. We are grateful to the engineers Francisco Evandro Albino, Leandro Melo and Marcelo Agudo Romão for the field works support.
References (58)
- et al.
A two-source time-integrated model for estimating surface fluxes using thermal infrared remote sensing
Remote Sens. Environ.
(1997) - et al.
Thermally derived evapotranspiration from the surface temperature initiated Closure (STIC) model improves cropland GPP estimates under dry conditions
Remote Sens. Environ.
(2022) - et al.
Assessing satellite-based basal crop coefficients for irrigated grapes (Vitis vinifera L.)
Agric. Water Manag., Amst.
(2010) - et al.
Combining a water balance model with evapotranspiration measurements to estimate total available soil water in irrigated and rainfed vineyards
Agric. Water Manag. Amst. V.
(2016) - et al.
Reflectance based crop coefficients REDUX: for operational evapotranspiration estimates in the age of high producing hybrid varieties
Agric. Water Manag., Amst.
(2017) - et al.
Water productivity and crop yield: a simplified remote sensing driven operational approach
Agric. For. Meteorol. Amst.
(2018) - et al.
Monitoring wheat phenology and irrigation in central Morocco: on the use of relationships between evapotranspiration, crop coefficients, leaf area index and remotely-sensed vegetation indices
Agric. Water Manag., V.
(2006) - et al.
Monitoring evapotranspiration o irrigated crops using crop coefficients derived from time series of satellite images II. Application on basin scale
Agric. Water Manag., Amst.
(2013) A soil‐adjusted vegetation index (SAVI)
Remote Sens. Environ.
(1988)- et al.
Development and validation of canopy reflectance-based crop coefficient for potato
Agric. Water Manag.
(2007)
Soil water content estimation using a remote sensing based hybrid evapotranspiration modeling approach
Adv. Water Resour., V.
A two‐source approach for estimating soil and vegetation energy fluxes in observations of directional radiometric surface‐temperature
Agric. For. Meteorol. Amst.
Correcting eddy-covariance flux underestimates over a grassland
Agric. For. Meteorol., Amst.
Field-scale mapping of evaporative stress indicators of crop yield: an application over Mead, NE, USA
Remote Sens. Environ.
Translating wind measurements from weather stations to agricultural crops
J. Hydrol. Eng.
Evaluation of a hybrid reflectance-Based crop coefficient and energy balance evapotranspiration model for irrigation management
ASABE
Crop coefficients derived from reflected canopy radiation ‐ a concept
Trans. ASAE
Acompanhamento da safra brasileira: Cana-de-açúcar, v.8 - Safra 2021/2022
Terc. Levant., Brasília
Varietal variation in growth, physiology and yield of sugarcane under two contrasting water regimes
Trop. Agric. Res., V.
Dimensioning the impact of irrigation on sugarcane yield in Brazil
Sugar Tech.
Assessing landscape scale heterogeneity in irrigation water use with remote sensing and in situ monitoring
Environ. Res. Lett..
Spatial evapotranspiration modelling interface (SETMI). In remote sensing and hydrology, editado por Neale, C.M.U.; Cosh, M.H
Sciences
Nutritional balance and production of sugarcane irrigated with treated wastewater through subsurface drip
Irrig. Sci. Berl.
Cited by (13)
Ensemble learning-based estimation of reference evapotranspiration (ET<inf>o</inf>)
2023, Internet of Things (Netherlands)Intelligent optimization of Reference Evapotranspiration (ETo) for precision irrigation
2023, Journal of Computational ScienceRemote sensing-based evapotranspiration modeling using geeSEBAL for sugarcane irrigation management in Brazil
2022, Agricultural Water ManagementCitation Excerpt :According to the Koppen classification (Peel et al., 2007), the climate of the region is defined as tropical type (Aw) with dry winter and rainy summer, with average annual precipitation of 1.242 mm, average annual reference evapotranspiration of 1536 mm, average solar radiation of 17 MJ m−2 day−1, average air temperature of 23º C, and average relative humidity of 62% (UNESP, 2019). The soil is classified as Typical Dystrophic Red Latosol (Santos et al., 2018), deep with sandy loam texture, more details about the soil's physical characteristics are in Bispo et al. (2022) and Table 1. The meteorological data used were collected from the automated weather station (Northwestern São Paulo Network - http://clima.feis.unesp.br) near the study site, and used to estimate the reference evapotranspiration (ETo) based on the FAO-Penman-Monteith method (Allen et al., 1998).
WINDS Model Simulation of Guayule Irrigation
2023, Water (Switzerland)