Evapotranspiration partitioning assessment using a machine-learning-based leaf area index and the two-source energy balance model with sUAV information.

Accurate quantification of the partitioning of evapotranspiration (ET) into transpiration and evaporation fluxes is necessary to understanding ecosystem interactions among carbon, water, and energy flux components. ET partitioning can also support the description of atmosphere and land interactions and provide unique insights into vegetation water status. Previous studies have identified leaf area index (LAI) estimation as a key descriptor of biomass conditions needed for the estimation of transpiration and evaporation. LAI estimation in clumped vegetation systems, such as vineyards and orchards, has proven challenging and is strongly related to crop phenological status and canopy management. In this study, a feature extraction model based on previous research was built to generate a total of 202 preliminary variables at a 3.6-by-3.6-meter-grid scale based on submeter-resolution information from a small Unmanned Aerial Vehicle (sUAV) in four commercial vineyards across California. Using these variables, a machine learning model called eXtreme Gradient Boosting (XGBoost) was successfully built for LAI estimation. The XGBoost built-in function requires only six variables relating to vegetation indices and temperature to produce high-accuracy LAI estimation for the vineyard. Using the six-variable XGBoost-based LAI map, two versions of the Two-Source Energy Balance (TSEB) model, TSEB-PT and TSEB-2T were used for energy balance and ET partitioning. Comparing these results with the Eddy-Covariance (EC) tower data, showed that TSEB-PT outperforms TSEB-2T on the estimation of sensible heat flux (within 13% relative error) and surface heat flux (within 34% relative error), while TSEB-2T outperforms TSEB-PT on the estimation of net radiation (within 14% relative error) and latent heat flux (within 2% relative error). For the mature vineyard (north block), TSEB-2T performs better than TSEB-PT in partitioning the canopy latent heat flux with 6.8% relative error and soil latent heat flux with 21.7% relative error; however, for the younger vineyard (south block), TSEB-PT performs better than TSEB-2T in partitioning the canopy latent heat flux with 11.7% relative error and soil latent heat flux with 39.3% relative error.

Incorporation of Unmanned Aerial Vehicle (UAV) Point Cloud Products into Remote Sensing Evapotranspiration Models.

In recent years, the deployment of satellites and unmanned aerial vehicles (UAVs) has led to production of enormous amounts of data and to novel data processing and analysis techniques for monitoring crop conditions. One overlooked data source amid these efforts, however, is incorporation of 3D information derived from multi-spectral imagery and photogrammetry algorithms into crop monitoring algorithms. Few studies and algorithms have taken advantage of 3D UAV information in monitoring and assessment of plant conditions. In this study, different aspects of UAV point cloud information for enhancing remote sensing evapotranspiration (ET) models, particularly the Two-Source Energy Balance Model (TSEB), over a commercial vineyard located in California are presented. Toward this end, an innovative algorithm called Vegetation Structural-Spectral Information eXtraction Algorithm (VSSIXA) has been developed. This algorithm is able to accurately estimate height, volume, surface area, and projected surface area of the plant canopy solely based on point cloud information. In addition to biomass information, it can add multi-spectral UAV information to point clouds and provide spectral-structural canopy properties. The biomass information is used to assess its relationship with in situ Leaf Area Index (LAI), which is a crucial input for ET models. In addition, instead of using nominal field values of plant parameters, spatial information of fractional cover, canopy height, and canopy width are input to the TSEB model. Therefore, the two main objectives for incorporating point cloud information into remote sensing ET models for this study are to (1) evaluate the possible improvement in the estimation of LAI and biomass parameters from point cloud information in order to create robust LAI maps at the model resolution and (2) assess the sensitivity of the TSEB model to using average/nominal values versus spatially-distributed canopy fractional cover, height, and width information derived from point cloud data. The proposed algorithm is tested on imagery from the Utah State University AggieAir sUAS Program as part of the ARS-USDA GRAPEX Project (Grape Remote sensing Atmospheric Profile and Evapotranspiration eXperiment) collected since 2014 over multiple vineyards located in California. The results indicate a robust relationship between in situ LAI measurements and estimated biomass parameters from the point cloud data, and improvement in the agreement between TSEB model output of ET with tower measurements when employing LAI and spatially-distributed canopy structure parameters derived from the point cloud data.

Estimation of Evapotranspiration and Energy Fluxes using a Deep-Learning based High-Resolution Emissivity Model and the Two-Source Energy Balance Model with sUAS information.

Surface temperature is necessary for the estimation of energy fluxes and evapotranspiration from satellites and airborne data sources. For example, the Two-Source Energy Balance (TSEB) model uses thermal information to quantify canopy and soil temperatures as well as their respective energy balance components. While surface (also called kinematic) temperature is desirable for energy balance analysis, obtaining this temperature is not straightforward due to a lack of spatially estimated narrowband (sensor-specific) and broadband emissivities of vegetation and soil, further complicated by spectral characteristics of the UAV thermal camera. This study presents an effort to spatially model narrowband and broadband emissivities for a microbolometer thermal camera at UAV information resolution (~0.15 m) based on Landsat and NASA HyTES information using a deep learning (DL) model. The DL model is calibrated using equivalent optical Landsat / UAV spectral information to spatially estimate narrowband emissivity values of vegetation and soil in the 7-14-nm range at UAV resolution. The resulting DL narrowband emissivity values were then used to estimate broadband emissivity based on a developed narrowband-broadband emissivity relationship using the MODIS UCSB Emissivity Library database. The narrowband and broadband emissivities were incorporated into the TSEB model to determine their impact on the estimation of instantaneous energy balance components against ground measurements. The proposed effort was applied to information collected by the Utah State University AggieAir small Unmanned Aerial Systems (sUAS) Program as part of the ARS-USDA GRAPEX Project (Grape Remote sensing Atmospheric Profile and Evapotranspiration eXperiment) over a vineyard located in Lodi, California. A comparison of resulting energy balance component estimates, with and without the inclusion of high-resolution narrowband and broadband emissivities, against eddy covariance (EC) measurements under different scenarios are presented and discussed.

The impact of shadows on partitioning of radiometric temperature to canopy and soil temperature based on the contextual two-source energy balance model (TSEB-2T).

Tests of the most recent version of the two-source energy balance model have demonstrated that canopy and soil temperatures can be retrieved from high-resolution thermal imagery captured by an unmanned aerial vehicle (UAV). This work has assumed a linear relationship between vegetation indices (VIs) and radiometric temperature in a square grid (i.e., 3.6 m × 3.6 m) that is coarser than the resolution of the imagery acquired by the UAV. In this method, with visible, near infrared (VNIR), and thermal bands available at the same high-resolution, a linear fit can be obtained over the pixels located in a grid, where the x-axis is a vegetation index (VI) and the y-axis is radiometric temperature. Next, with an accurate VI threshold that separates soil and vegetation pixels from one another, the corresponding soil and vegetation temperatures can be extracted from the linear equation. Although this method is simpler than other approaches, such as TSEB with Priestly-Taylor (TSEB-PT), it could be sensitive to VIs and the parameters that affect VIs, such as shadows. Recent studies have revealed that, on average, the values of VIs, such as normalized difference vegetation index (NDVI) and leaf area index (LAI), that are located in sunlit areas are greater than those in shaded areas. This means that involving or compensating for shadows will affect the linear relationship parameters (slope and bias) between radiometric temperature and VI, as well as thresholds that separate soil and vegetation pixels. This study evaluates the impact of shadows on the retrieval of canopy and soil temperature data from four UAV images before and after applying shadow compensation techniques. The retrieved temperatures, using the TSEB-2T approach, both before and after shadow correction, are compared to the average temperature values for both soil and canopy in each grid. The imagery was acquired by the Utah State University AggieAir UAV system over a commercial vineyard located in California as part of the USDA Agricultural Research Service Grape Remote sensing Atmospheric Profile and Evapotranspiration Experiment (GRAPEX) Program during 2014 to 2016. The results of this study show when it is necessary to employ shadow compensation methods to retrieve vegetation and soil temperature directly.

Validation of digital surface models (DSMs) retrieved from unmanned aerial vehicle (UAV) point clouds using geometrical information from shadows.

Theoretically, the appearance of shadows in aerial imagery is not desirable for researchers because it leads to errors in object classification and bias in the calculation of indices. In contrast, shadows contain useful geometrical information about the objects blocking the light. Several studies have focused on estimation of building heights in urban areas using the length of shadows. This type of information can be used to predict the population of a region, water demand, etc., in urban areas. With the emergence of unmanned aerial vehicles (UAVs) and the availability of high- to super-high-resolution imagery, the important questions relating to shadows have received more attention. Three-dimensional imagery generated using UAV-based photogrammetric techniques can be very useful, particularly in agricultural applications such as in the development of an empirical equation between biomass or yield and the geometrical information of canopies or crops. However, evaluating the accuracy of the canopy or crop height requires labor-intensive efforts. In contrast, the geometrical relationship between the length of the shadows and the crop or canopy height can be inversely solved using the shadow length measured. In this study, object heights retrieved from UAV point clouds are validated using the geometrical shadow information retrieved from three sets of high-resolution imagery captured by Utah State University's AggieAir UAV system. These flights were conducted in 2014 and 2015 over a commercial vineyard located in California for the USDA Agricultural Research Service Grape Remote sensing Atmospheric Profile and Evapotranspiration Experiment (GRAPEX) Program. The results showed that, although this approach could be computationally expensive, it is faster than fieldwork and does not require an expensive and accurate instrument such as a real-time kinematic (RTK) GPS.

Estimation of surface thermal emissivity in a vineyard for UAV microbolometer thermal cameras using NASA HyTES hyperspectral thermal, Landsat and AggieAir optical data.

Microbolometer thermal cameras in UAVs and manned aircraft allow for the acquisition of high-resolution temperature data, which, along with optical reflectance, contributes to monitoring and modeling of agricultural and natural environments. Furthermore, these temperature measurements have facilitated the development of advanced models of crop water stress and evapotranspiration in precision agriculture and heat fluxes exchanges in small river streams and corridors. Microbolometer cameras capture thermal information at blackbody or radiometric settings (narrowband emissivity equates to unity). While it is customary that the modeler uses assumed emissivity values (e.g. 0.99-0.96 for agricultural and environmental settings); some applications (e.g. Vegetation Health Index), and complex models such as energy balance-based models (e.g. evapotranspiration) could benefit from spatial estimates of surface emissivity for true or kinetic temperature mapping. In that regard, this work presents an analysis of the spectral characteristics of a microbolometer camera with regard to emissivity, along with a methodology to infer thermal emissivity spatially based on the spectral characteristics of the microbolometer camera. For this work, the MODIS UCBS Emissivity Library, NASA HyTES hyperspectral emissivity, Landsat, and Utah State University AggieAir UAV surface reflectance products are employed. The methodology is applied to a commercial vineyard agricultural setting located in Lodi, California, where HyTES, Landsat, and AggieAir UAV spatial data were collected in the 2014 growing season. Assessment of the microbolometer spectral response with regards to emissivity and emissivity modeling performance for the area of study are presented and discussed.

Real-time reservoir operation using data mining techniques.

The optimal operation of hydropower reservoirs is essential for the planning and efficient management of water resources and the production of hydroelectric energy. Various techniques such as the genetic algorithm (GA), artificial neural networks (ANN), support vector machine (SVM), and dynamic programming (DP) have been employed to calculate reservoir operation rules. This paper implements the data mining techniques SVM and ANN to calculate the optimal release rule of hydropower reservoirs under "forecasting" and "non-forecasting" scenarios. The employment of data mining techniques accounting for data uncertainty to calculate optimal hydropower reservoir operation is novel in the field of water resource systems analysis. The optimal operation of the Karoon 3 reservoir, Iran, serves as a test of the proposed methodology. The upstream streamflow, storage records, and several lagged variables are model inputs. Data obtained from solving the reservoir optimization problem with nonlinear programming (NLP) are applied to train (calibrate) the SVM, and ANN, SVM, and ANN are executed in the "non-forecasting" scenario based on all inputs along with their time-lagged variables. In contrast, current parameters are removed from the set of inputs in the "forecasting" scenario. The results of the SVM model are compared with those of the ANN model with the correlation coefficient (R), the mean error (ME), and the root mean square error (RMSE). This paper's results indicate performance of the SVM is better than that of the ANN model by 1.5%, 400%, and 10% with respect to the R, ME, and RMSE diagnostic statistics, respectively. In addition, SVM and ANN overcome data uncertainty ("forecasting" scenario) to produce optimal reservoir operation.

Implications of sensor inconsistencies and remote sensing error in the use of small unmanned aerial systems for generation of information products for agricultural management.

Small, unmanned aerial systems (sUAS) for remote sensing represent a relatively new and growing technology to support decisions for agricultural operations. The size and power limitations of these systems present challenges for the weight, size, and capability of the sensors that can be carried, as well as the geographical coverage that is possible. These factors, together with a lack of standards for sensor technology, its deployment, and data analysis, lead to uncertainties in data quality that can be difficult to detect or characterize. These, in turn, limit comparability between data from different sources and, more importantly, imply limits on the analyses that can be accomplished with the data that are acquired with sUAS. This paper offers a simple statistical examination of the implications toward information products of an array of sensor data uncertainty issues. The analysis relies upon high-resolution data collected in 2016 over a commercial vineyard, located near Lodi, California, for the USD A Agricultural Research Service Grape Remote sensing Atmospheric Profile and Evapotranspiration experiment (GRAPEX) Program. A Monte Carlo analysis is offered of how uncertainty in sensor spectral response and/or orthorectification accuracy can affect the estimation of information products of potential interest to growers, as illustrated in the form of common vegetation indices.

Behavior of vegetation/soil indices in shaded and sunlit pixels and evaluation of different shadow compensation methods using UAV high-resolution imagery over vineyards.

In high-resolution imagery, shadows may cause problems in object segmentation and recognition due to their low reflectance. For instance, the spectral reflectance of shadows and water are similar, particularly in the visible band. In precision agriculture, the vegetation condition in terms of plant water use, plant water stress, and chlorophyll content can be estimated using vegetation indices. Normalized Difference Vegetation Index (NDVI), Leaf Area Index (LAI) and Enhanced Vegetation Index (EVI) are widely used vegetation indices for characterizing the condition of the vegetation. In addition, many soil indices have been developed for describing soil characteristics, such as Soil-Adjusted Vegetation Index (SAVI). However, shadows can have an influence on the performance of these vegetation and soil indices.Moreover, enhancing spatial resolution heightens the impact of shadows in the imagery. In this study, the behavior of vegetation and soil indices are evaluated using four sets of high-resolution imagery captured by the Utah State University AggieAir unmanned aerial vehicle (UAV) system. These indices were obtained from flights conducted in 2014, 2015, and 2016 over a commercial vineyard located in California for the USDA Agricultural Research Service Grape Remote sensing Atmospheric Profile and Evapotranspiration Experiment (GRAPEX) Program. Different shadow restoration methods are used to alleviate the impact of shadows in information products that might be developed from the high-resolution imagery. The histogram pattern of vegetation and soil indices before and after shadow compensation, are compared using sanalysi of variance (ANOVA). The results of this study indicate how shadows can affect the vegetation/soil indices and whether shadow compensation methods are able to remove the statistical difference between sunlit and shadowed vegetation/soil indices.

Assessment of different methods for shadow detection in high-resolution optical imagery and evaluation of shadow impact on calculation of NDVI, and evapotranspiration.

Significant efforts have been made recently in the application of high-resolution remote sensing imagery (i.e., sub-meter) captured by unmanned aerial vehicles (UAVs) for precision agricultural applications for high-value crops such as wine grapes. However, at such high resolution, shadows will appear in the optical imagery effectively reducing the reflectance and emission signal received by imaging sensors. To date, research that evaluates procedures to identify the occurrence of shadows in imagery produced by UAVs is limited. In this study, the performance of four different shadow detection methods used in satellite imagery was evaluated for high-resolution UAV imagery collected over a California vineyard during the Grape Remote sensing Atmospheric Profile and Evapotranspiration eXperiment (GRAPEX) field campaigns. The performance of the shadow detection methods was compared and impacts of shadowed areas on the normalized difference vegetation index (NDVI) and estimated evapotranspiration (ET) using the Two-Source Energy Balance (TSEB) model are presented. The results indicated that two of the shadow detection methods, the supervised classification and index-based methods, had better performance than two other methods. Furthermore, assessment of shadowed pixels in the vine canopy led to significant differences in the calculated NDVI and ET in areas affected by shadows in the high-resolution imagery. Shadows are shown to have the greatest impact on modeled soil heat flux, while net radiation and sensible heat flux are less affected. Shadows also have an impact on the modeled Bowen ratio (ratio of sensible to latent heat) which can be used as an indicator of vine stress level.