Summit Abstracts
Drone based Multispectral and Thermal Imagery for Geospatial Water Use Mapping of Irrigated Field Crops
Abhilash Chandel, Washington State University
Geospatial crop water use mapping is critical for field-scale irrigation management. Landsat 7/8 satellite imagery with a widely adopted METRIC energy balance model (LM approach) estimates accurate evapotranspiration (ET) but limits field-scale spatiotemporal mapping (30 m/pixel, 16 days). A study was therefore conducted to map actual ET of commercially grown irrigated-field crops (spearmint, potato, and alfalfa) at very high-resolution (7 cm/pixel). Six small unmanned aerial system (UAS)-based multispectral and thermal infrared imagery campaigns were conducted (two for each crop) at the same time as the Landsat 7/8 overpass. Three variants of METRIC model were used to process the UAS imagery: UAS-METRIC-1, -2, and -3 (UASM-1, -2, and -3) and outputs were compared with the standard LM approach. UASM-2 had the highest similarity with the LM approach (daily ET departures: 2–14%, Pearson correlation coefficient = 0.91). UASM approaches (Coefficient of variation, CV: 6.7–24.3%) however outperformed the LM approach (CV: 2.1–11.2%) in mapping spatial ET variations. On-demand UAS imagery may thus help in deriving high resolution site-specific ET maps for growers to aid in timely crop water management.
Intelligent sprayer for pesticide applications in modern orchard systems
Anura Rathnayake, Washington State University
An intelligent sprayer retrofit system developed by the USDA-ARS is being adapted to commercial spray applications. It uses Light Detection and Ranging (LiDAR) sensing technology to assess tree canopy structure in combination with GPS to track sprayer ground speed and position. Such data is processed by an onboard computer which calculates output spray rates for pertinent trees. These spray rates are then signaled to the embedded flow controller to proportionally control each nozzle independently through solenoid valves using pulse width modulation. Through such decision making, the system controls droplet size and over-spray reducing chemical usage, pesticide drift and improving efficient spray application. For optimized operation, the retrofit system enables the user to input spray rate (ounces of liquid to treat one cubic foot of canopy), based on the canopy structure or architecture. Such user inputs are guided by seemingly arbitrary assessments which may affect the actual spray outcomes i.e. deposition, coverage and efficacy. We therefore aim to conduct experimental studies to develop user guidance on input spray rates for effective chemical applications in modern orchard systems of the Washington State.
Investigating artificial hot and cold reference surfaces for estimating water use with thermal images
Behnaz Molaei, Washington State University
Thermal imaging can be used to monitor crop water status using a crop water stress index (CWSI) or evapotranspiration (ET) using water-energy balance models like SEBAL or METRIC. Temperature references from the images representing ET (METRIC) or relative transpiration (CWSI) with no water limitation (Tcold) and with extreme water limitation (Thot), where ET or transpiration are negligible, are needed. Artificial hot and cold reference surfaces might simplify the processing of UAV-based thermal images that are often not large enough to include representative and uniform Tcold and Thot surfaces. A study was set up at WSU-IAREC, near an AgWeatherNet weather station to find suitable surfaces for use as artificial hot and cold references and compare these surfaces with the temperatures that were obtained from theoretical energy balance approaches. Green healthy ryegrass was planted in the boxes and they were fully irrigated to use as a cold reference surface and a box of dry grass was used as the non-transpiring hot reference surface during this two-year (2017-2018) study. Thirteen different miscellaneous hot and cold surfaces were placed close to the grass boxes. Thermal and RGB cameras were mounted on a ground vehicle with a retract
Strawberry Crops Segmentation System using Computer Vision Strategies
Camilo Pardo-Beainy, Santo Tomás University
Traditionally, the analysis of strawberry crop information is performed without the use of technological elements, and in many cases, to diagnose a problem in the plant, it is necessary to perform destructive laboratory tests, such as foliar analysis and detection of diseases in plants and fruits. Also, in large areas of crops, these analyzes can only be applied to a few samples of crop plants; omitting so much of the information that may be relevant for production.
This project is primarily focused in to realize monitoring processes using computer vision and artificial intelligence techniques, to track characteristics of fruit crops and yield evolution. To carry out this objective, has been necessary to test different techniques of segmentation of the images of the crop that allow the identification of strawberry plants, separating the plant from other unwanted elements. Color analysis techniques, unsupervised segmentation through the application of Clustering and the use of thresholds are used to perform the segmentation procedure. Subsequently, characteristic extraction systems are implemented to carry out procedures to identify homogeneous fruit maturity and crop yield projections.
The effect of the foliar enrichment and herbicides on maize and associated weeds irrigated with drainage water
Ibrahim Mosaad dribrahim, Agriculture Research Center
A two-year field experiment was conducted during summer seasons of 2013 and 2014, which were irrigated by drainage water which belong to salinity class (C3S1 to C4S2), to study the effect of the foliar enrichment namely (Anti-stress) and weed management treatments (some pre and post-emergence herbicides and two-hand hoeing) on maize growth, yield, yield components and chemical composition of maize grains and associated weeds (Portulaca oleracea, Amaranthus retroflexus and Echinochloa colonum). The results illustrated that application of the foliar enrichment enhanced the dry weight of weeds and increased maize growth characters, yield and yield components and total crude protein and total oil percentage of grain maize, as compared with untreated treatment. All weed management treatments caused a significant reduction in total dry weight of weeds at 60 and 80 days after sowing in both seasons. Two-hand hoeing treatment exerted the highest decrease in total dry weight of weeds followed by metribuzin, oxadiagyl, fluroxypyr and bentazon, respectively at 60 and 80 days after sowing compared with other weed management treatments.
A biogeochemical-economic model for the valuation of cover crops ecosystem services under climate change
Karen Moran-Rivera, University of Hawai’i at Manoa
Despite all the well-known cover crop (CC) benefits, widespread adoption in the Midwestern U.S. is still low. There continues to be a debate about whether adopting cover crop (CC) is privately optimal for farmers and how climate change might affect the private incentives to adopt. We developed a biogeochemical-economic model that estimates the ecosystem service benefits provided by CC under different climate scenarios on a corn-soybean farm and contrasts them with CC costs over 10 years. We used the DeNitrification-DeComposition (DNDC) model as the ecological production function in the biogeochemical-economic model. The biogeochemical-economic model simulation results suggest that under most climate scenarios CC adoption does not generate a sizable difference in farm net present values (NPVs). However, under frequent extreme droughts, adopting CC increases a farm’s NPV by 15%. This difference is explained by higher corn yields in the CC treatment, where corn yields were 15% higher under frequent extreme droughts. DNDC simulation results show that this yield increase is due to an increase in three ecosystem services in the CC system: improved soil water storage, soil organic matter accumulation, and N retention.
Internet of Things enabled crop physiology sensing unit for apple sunburn management
Rakesh Ranjan, Washington state university
Heat and light instigated abiotic stresses cause considerable crop loss. Such stressors escalate the fruit surface temperature (FST) and prolonged exposures above critical FST may result in apple sunburn. This study emphasizes on the development of Internet of Things (IoT) enabled crop physiology sensing (CPS) unit for noninvasive sunburn monitoring. Each of the CPS node integrates a thermal-RGB sensor and a microclimate sensor with an edge computing device (Pi V3B, Raspberry Foundation Inc., UK). The CPS unit was programmed to processes all the data onboard the CPS unit and save the estimated FSTs on cloud server and remote host computer in real time. This study reports the operational performance of CPS unit tested in year 2019 for apple cultivars cv. Honeycrisp and Cosmic Crisp. The FST data was collected for three consecutive days between 12–5 pm at 5-minute intervals. The result of the study indicates that imagery derived apple FST had a strong correlation (r = 0.8) with ground truth measured FST. Overall, developed CPS unit performed reliably for sunburn monitoring. Additionally, this presentation will report the season long sunburn monitoring and management efforts made in production season 2020 for Cosmic Crisp apple.