Source: WASHINGTON STATE UNIVERSITY submitted to NRP
CPS: MEDIUM: FIELD-SPECIFIC WEATHER-DRIVEN AUTOMATED FROST MITIGATION OF SPECIALTY CROPS
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
ACTIVE
Funding Source
AFRI COMPETITIVE GRANT
Reporting Frequency
Annual
Accession No.
1025792
Grant No.
2021-67021-34336
Cumulative Award Amt.
$1,009,522.00
Proposal No.
2020-11346
Multistate No.
(N/A)
Project Start Date
Feb 15, 2021
Project End Date
Feb 14, 2026
Grant Year
2021
Program Code
[A7302]- Cyber-Physical Systems
Recipient Organization
WASHINGTON STATE UNIVERSITY
240 FRENCH ADMINISTRATION BLDG
PULLMAN,WA 99164-0001
Performing Department
AgWeatherNet Prograam
Non Technical Summary
Frost events present a significant threat to perennial crop producers due both to potential bud damage and mitigation costs. Growers struggle to optimize the utilization of three primary active mitigation methods--heaters, wind machines, and over- and under-tree sprinkler irrigation--due to a reliance on regional weather forecasts, rules-of-thumb, and heuristic decision-making. Our overall project objective is to develop an intelligent frost mitigation control actuation system, based on micrometeorological monitoring and site-specific, mitigation-informed weather and bud temperature forecasts, that reduces both crop damage and mitigation costs. Washington State is the top producer of sweet cherries and blueberries--both vulnerable to frost damage and the focus of this project.Toward the overall project objective, we will pursue three specific aims: Aim #1. Integrate surface?and?aerial?meteorological?observations into?field-specific,?short-term?forecasts?of weather variables relevant to frost mitigation. Aim #2. Develop localized weather data-driven intelligent crop loss management system through real-time actuation of either-or combinations of active frost mitigation techniques, i.e. wind machines (frost fans), over-tree and under-tress fixed spray systems. Aim #3. Assess grower?evaluation/validation of decision aid tools and prototype performance.
Animal Health Component
60%
Research Effort Categories
Basic
10%
Applied
60%
Developmental
30%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
4041112202030%
4041120202030%
1320499207040%
Knowledge Area
132 - Weather and Climate; 404 - Instrumentation and Control Systems;

Subject Of Investigation
1120 - Blueberry; 1112 - Cherry; 0499 - Atmosphere, general/other;

Field Of Science
2070 - Meteorology and climatology; 2020 - Engineering;
Goals / Objectives
Our overall project objective is to develop an intelligent frost mitigation control actuation system, based on micrometeorological monitoring and site-specific, mitigation-informed weather and bud temperature forecasts, that reduces both crop damage and mitigation costs. Washington State is the top producer of sweet cherries and blueberries--both vulnerable to frost damage and the focus of this project.Toward the overall project objective, we will pursue three specific aims: Aim #1. Integrate surface and aerial meteorological observations into field-specific, short-term forecasts of weather variables relevant to frost mitigation. Aim #2. Develop localized weather data-driven intelligent crop loss management system through real-time actuation of either-or combinations of active frost mitigation techniques, i.e. wind machines (frost fans), over-tree and under-tress fixed spray systems. Aim #3. Assess grower evaluation/validation of decision aid tools and prototype performance.
Project Methods
Aim 1. The four tasks in this aim are predicated on the hypothesis that site-specific forecasts can be improved by postprocessing weather model output with observations, a technique used at airports since the 1970s and has recently been improved through advances in artificial intelligence/machine learning. This project further improves this technique by adding a module to forecast site-specific weather (Task 1), measuring the effect of wind machines and irrigation on bud temperature (Tasks 2-3) and using that information to develop advisory forecasts (Task 4).Tasks 2-3 focus on the collection of the bud temperature forecasts based on the weather forecasts from Task 1 and the predicted impact of wind machines and irrigation. We will collect inside and outside of the orchard profiles of temperature, dew point, and wind in the lowest ~150 m of the atmosphere to determine the presence of an inversion as well as the impact of fans on the inversion. Task 4 combines data from Tasks 1-3, to create 4 sets of hourly bud temperature forecasts based on the expected impact of different frost mitigation practices.Aim 2. The project team will work toward integrating site-specific weather data forecasts (Aim #1 output) and related sensing data to timely actuate active frost mitigation techniques. We will develop a field deployable intelligent sensing and control system (Task 1) that learns from site-specific weather inversion forecasts, real-time weather, bud temperature data and crop specific grower knowledge to actuate individual or a combination of mitigation techniques. In terms of frost mitigation, the novel shower-down fixed spray systems researched in PI-Khot's precision agriculture group, to apply crop chemicals in perennial specialty crops, will be engineered as overtree irrgation based frost mitigation and will be actuated either individually or in combination with undertree irrigation and commercial wind machines. We will integrate and field deploy a prototype monitoring system with active frost mitigation technologies (Task 2) and refine the functionality for reliability of actuation through field experiments (Task 3).Task 1: A proposed IoT-enabled frost monitoring network will monitor the in-field weather, tree canopy & bud physiology to provide localized and spatial data inputs. The plan is to create a network of interconnected sensing nodes and deploy them in the field to obtain real-time data directly at the field block level. This step will be followed by performing data mining with AWN bud temperature forecasts for need-based, real-time actuation of frost mitigation techniques.Task 2: A fixed spray delivery system will be utilized to apply the water mist to the canopies for preventing the frost related bud injuries in sweet cherry and highbush blueberry. Such micro-emitter and overhead fogger/mister configured SSCDS will enable a precise canopy misting for efficient and economical frost management with up to 80% saving in the water compared to coarse droplet producing conventional overhead nozzles. The nozzles of the in-field undertree irrigation system will be replaced by a low volume micro-emitter (with swivel deflector) to enhance system efficacy.Task 3: In Year 2, an automated frost management system will be tested in one of the frost susceptible cultivars of highbush blueberry ('Duke or 'Draper') and sweet cheery ('Bing' or 'Skeena') planted in an experimental block. Similarly, an optimized prototype will be deployed in Year-3, with 300 ft × 4 irrigation lines as a test plot × 3 replicate plots for both commodities. The sensor network will be deployed, in consultation with growers, to provide real-time data on temperature variation, which will be both an input to the decision making with the intelligent controller and an output tracking the impact of mitigation technologies. The flowmeters will be installed in irrigation lines to estimate the water savings/day/week/season and savings in pumping energy. Such data will be contrasted against the conventional system with continuous operation and manual wind machine operations. Efficacy of the integrated system will be evaluated by monitoring bud temperature data, through existing sensing nodes and by aerial flights of small UAS integrated with radiometric calibrated thermal imager.Aim 3 Efforts: The extension education component will be focused on documenting the effectiveness of proposed frost mitigation technologies relative to current practices. Regional extension specialist Gwen Hoheisel will lead these efforts along with active participation by the research team. Task 1 will include soliciting grower input and documenting ongoing frost mitigation practices through involved participation in at least three annual winter advisory meetings. Selected participants will be invited to summer focus group meetings to review the frost season management decision making. We will also frequently check in with growers during cold events to understand which mitigation practices were deployed and why. From year 2, the team will identify at least 6 beta testers to be provided with prototype decision forecasts that the testers can compare with their own decision making based on existing forecasts (Task 2).Efforts will be evaluated through targeted surveys of all participating growers, focus group participants and beta testers.

Progress 02/15/24 to 02/14/25

Outputs
Target Audience:Scientific community (Meteorology, earth science, horticulturists, and Agricultural Engineering); Growers, crop consultants and farm managers. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Graduate students presented the technology specific results in local grower meetings (e.g., 120th Washington State Tree Fruit Association hosted Annual Meeting and NW Hort. Expo.), and at American Society of Agricultural and Biological Engineers AIM 2024. Through this meeting participation, students received direct feedback and networked with fellow researchers. Students also demonstrated the technology in Smart Orchard Field Day 2024 and to K-12 students and teachers during their visit to WSU PrecisionAg Lab. The PhD student translated research into teaching by presenting dronesonde integration and frost mitigation use case as part of the department course BYSE 552: Unmanned Aerial Systems and Precision Agriculture (2 credits). How have the results been disseminated to communities of interest?The results were shared within the research community and grower stakeholders via presentations at ASABE AIM 2024, local grower meetings and as media/news articles and interviews. The models to predict all critical weather variables for frost mitigation, results of dronesonde, and components of localized intelligent crop loss management system with preliminary studies on radiative frost sensor were presented during local grower meetings. In 2022, AgWeatherNet (AWN) conducted a Weather School where the recent findings of this experiment were shared. Videos of the presentations were edited in 2023, posted on AWN and CAHNRS you-tube channel, as well as shared through the following newsletters: WSU Viticulture, WSU Potatoes, Small Fruit Update. In 2024, the videos will be released monthly to the WSU tree fruit newsletter. These subscriptions collectively reach more than 3000 producers. The videos include the following: Climate: Why Does It Matter Meteorology Basics: Why They Matter AgWeatherNet Overview Weather Sensing: Principals & Tiers Soil Water Potential AgWeatherNet Web Portal Weather Sensing: Adding Private Stations to AWN (Hardware & Software) Weather Data Quality Assurance Cold Hardiness of Grapes, Blueberries and Cherries Private Weather Station Options (4 Companies) Forecasting Station-Specific Weather DAS - Potatoes DAS Overview What do you plan to do during the next reporting period to accomplish the goals?Aim# 1. Predictions of out-of-orchard reference station weather are useful, however, since weather conditions in the orchard can vary substantially from the cleared, flat, reference site, we have started quantifying both to understand how the reference station weather is related to In-orchard conditions. These efforts will continue in 2025 season. We plan to incorporate of Out-of-orchard to In-orchard weather prediction model onto the existing model framework to translate out-of-orchard forecasts to in-orchard forecasts using data collected in Spring 2023, 2024 and 2025. Based on these offsets, short-term orchard specific nowcasting model will be developed for temperature and inversion forecasts. Aim #2.Frost events will be monitored in sweet cherry orchards and blueberry farm for Spring 2025 using a wireless sensing network, similar to previous year. The bud tissue temperature prediction model will be further validated using the data collected at multiple locations and heat maps will be generated for a given orchard block. Using air temperature data inside the orchard at 1.5 and 9 m, a thorough understanding on the inversion strength and wind machine efficiency will be developed. Offsets will be developed for the orchard blocks using reference open weather station data. An algorithm for stratified sensor deployment using drone thermal infrared imager will be revised. Orchard specific weather and bud tissue temperature nowcasting model will be improved using 2025 field season data and compared with reference open weather station forecasts. Overall, a decision support system comprising bud tissue temperature prediction and weather nowcasting model will be integrated into a publicly assessable dashboard to notify orchard managers indicating potential frost events in sweet cherry orchards. Aim #3. We will continue to engage with grower cooperators and share project knowledge at industry annual meetings and WSU-organized field days. A factsheet summarizing weather stations, their installation and use, will be disseminated to the tree fruit and grape industries. Webinars have been, and will continue to be, conducted through AgWeatherNet and Extension. Topics for these webinars include, among others: inversion forecasting systems, cold hardiness models, heat stress, weather-guided irrigation management. These webinars are usually recorded, and a link is posted on the AWN website. Two newsletter articles regarding frost management have been published and disseminated to the tree fruit and grape industries in 2024. In 2025, we plan to publish two extension articles on wireless sensor network for abiotic stress monitoring and use of station-specific inversion forecasts to manage frost.

Impacts
What was accomplished under these goals? Aim #1. Continuing the efforts from 2023, to realize field-specific short-term weather forecasts, weather stations were deployed inside the orchard (3 sites) to collect year-long air temperature, relative humidity, wind speed, and inversion strength (1.5 and 9 m AGL) data. Data collection will continue through Spring 2025. Such data is being utilized to understand the influence of micro-climate on the weather variables and to develop corrective offsets for reference open-field weather stations to represent the micro-climate more accurately. This data will also be used to re-train AWN weather station specific forecast models to realize short-term site-specific nowcasts. AgWeatherNet has already developed and operationalized the 10-day station specific forecast models through another project (WSDA-SCBG). These models have been evaluated against true observations across six forecast ranges, i.e., next-hour, nowcast (6 h), short-range (24 h), medium-range (72 h), long-range (120 h), and extended-range (240 h), with forecast errors categorized into different ranges. In terms of air temperature, for the next-hour forecast, 81% (of 169 open-field stations) stations had forecast errors below 1°C and 99% of the stations had forecast error below 3°C. Moreover, AWN has also operationalized inversion forecasts at mesonet tower stations with errors less than 1°C error over the next 24 hrs., essential for operating wind machines. Our on-going efforts through this project focus on using these station specific models, as a base, to develop orchard site-specific nowcasts and validate those using 2025 season data. Aim # 2.A wireless network capable of collecting a wide range of environmental data was deployed in three sweet cherry orchards and a blueberry farm. A stratified deployment method was used to monitor air and bud tissue temperature in the cold spots identified using a small UAS integrated thermal infrared imager. Localized air and bud tissue temperature were monitored during frost events in Spring of 2024. Similar to prior seasons, statistical results suggested the need for operational changes in current frost mitigation strategies. Using the Spring 2022 and 2023 season data, a bud tissue temperature prediction model has been developed. This model uses localized weather and radiative frost sensor data inputs and serves as a replacement to use of either air temperature or approach of thermocouple insertion into buds. Both sensing approaches are often inaccurate and problematic to use over extended time periods (i.e., through progression of bud growth stages). The developed model will be further validated in Spring 2025 season. To realize technology transition to growers, user-friendly web interface has been developed, as a tie-in to the existing field deployed wireless network (on three sites) and hosted on AgWeatherNet ecosystem https://weather.wsu.edu/ for real time frost conditions monitoring and management by our grower cooperators. For sprinkler-based frost mitigation in blueberry crops, a pilot experiment was conducted on potted blueberries (cv. Duke) using multiple micro-emitter configurations applicable for both frost and pest management. The system was automated using air temperature thresholds and the emitters run continuously until air temperature increases above set threshold. Water flow meters were installed to quantify water for the micro-emitters. Failed experiment was conducted in Spring 2024, as season never resulted in radiative frost conditions. Aim #3.Our team has collected field data and tested the developed system prototypes on grower cooperator sites. Prior to the start of the experiments, the team met with early adopters of technologies and cooperators on cold hardiness projects. Their feedback on project objectives and design was incorporated in the experiments. For two years, we have conducted an annual 'walk-in' clinic for growers to bring fruit buds to learn how to assess damage and use models. We have conducted multiple webinars on cold hardiness models (3), inversions (1), and measuring bud temperature (1). Lastly, a presentation at the WA Small Fruit Conference 2024 on the use of cold hardiness models was followed by a circle group discussion with producers on ways to improve the interface and industry awareness of how to use decision tools better. The results of those discussions have been shared with the WSU AgWeatherNet team for improvements of decision tools. Early spring field days are planned for the 2024-2025 frost season to showcase results and decision tools developed through this project. Surveys and follow-up focus interviews will be conducted at these field days.

Publications

  • Type: Theses/Dissertations Status: Published Year Published: 2024 Citation: Karisma Yumnam. December 2023. MS Thesis. Title: Development of localized in-orchard weather prediction model for apple crop management. Washington State University
  • Type: Other Journal Articles Status: Submitted Year Published: 2024 Citation: Yumnam, K., Cann, M. D., Khot*, L. R., Brown, D. J., and Kalcsits, L. 2024. Quantifying modern orchard systems and management effects on weather variables. Journal of Applied Meteorology and Climatology, JAMC-D-24-0154
  • Type: Other Journal Articles Status: Under Review Year Published: 2025 Citation: Gorthi S., Khot, L. R., and Hoheisel G.-A. 2025. Assessing direct and indirect methods to monitor bud tissue temperature for effective spring frost management in Sweet Cherry (Prunus avium cv. Chelan). Scientia Horticulture (Internal Review).
  • Type: Other Status: Published Year Published: 2024 Citation: Gorthi, S. Hoheisel, G. A. and Khot*, L. R., 2024. Is it important to monitor bud tissue temperature for spring frost management in sweet cherry? Washington State University  Fruit Matters, February 1, 2024.
  • Type: Other Status: Published Year Published: 2024 Citation: Gorthi*, S. Hoheisel, G. A. and Khot, L. R., 2024. Weather with AgWeatherNet: Air and bud tissue temperature sensing for frost mitigation in sweet cherry and blueberry [Virtual meeting]. Washington State University AgWeatherNet.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Gorthi*, S. Khot, L.R., 2024. LoRaWAN-Powered Wireless Sensing Network: Precision Abiotic Stress Management in Specialty Crops. Presented at the 120th Annual Meeting and NW Hort Expo, Washington State Tree Fruit Association, Yakima Convention Center, Yakima, WA. December 911, 2024 (Oral Presentation). Time: 5 min. Participants ~150.
  • Type: Other Journal Articles Status: Published Year Published: 2024 Citation: Amogi, B., Khot*, L. R., and Hill, L. (2024). Temperature inversion forecasts available for AgWeatherNet Mesonet tower stations. Washington State University  Fruit Matters, April 7, 2024.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Gorthi, S., Dhakar, R.*, Yumnam, K., Amogi, B., Khot, L. R., Hoheisel, G. A. 2024. Dronesonde and Intelligent Sensing Network for Frost Management. S1069: Research and Extension for Unmanned Aircraft Systems (UAS) Applications in U.S. Agriculture and Natural, Multistate Research Project Meeting, Montana State University, Bozeman, MT. June 13-15, 2024. (Poster Presentation).
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Gorthi, S.*, Dhakar, R., Yumnam, K., Amogi, B., Khot, L. R., Hoheisel, G. A. Intelligent wireless sensing network for effective frost management in sweet cherry (cv. Chelan) orchards. Honouring Undergraduate and Graduate Scholars Symposium, Heritage University, Toppenish, WA. April 5, 2024. Time: 60 min. Participants: ~75.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Amogi B. R., Cann M. D., Khot L. R. 2024. Station-specific weather and inversion forecast model development and verification for abiotic stress management in fruit crops. ASABE 2024. Annual International Meeting, Anaheim, CA. July 28-31, 2024. (Poster Presentation).
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Gorthi, S.*, Khot, L. R., Hoheisel, G. A. 2024. Sensor-fusion based sweet cherry (cv. Chelan) bud tissue temperature prediction model for effective spring frost management. ASABE 2024 Annual International Meeting, Anaheim, CA. July 28-31, 2024. (Poster Presentation).
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Amogi, B. R., Cann, M. D., Hill, S., Khot, L. (2024, December 911). Utilizing inversion forecast from AgWeatherNet for operating wind machines during early spring frost season. Presented at the 120th Annual Meeting and NW Hort Expo, Washington State Tree Fruit Association, Yakima Convention Center, Yakima, WA.
  • Type: Websites Status: Published Year Published: 2024 Citation: Prengaman, K. 2024. Precision frost protection research project looks at inversion events with weather towers and drones to build better models for growers. https://www.goodfruit.com/inversion-conditions-help-inform-frost-forecasts/
  • Type: Other Journal Articles Status: Accepted Year Published: 2025 Citation: Yumnam, K., Cann, M. D., Khot*, L. R., Brown, D. J., and Kalcsits, L. (2024). Quantifying modern orchard systems and management effects on weather variables. Journal of Applied Meteorology and Climatology, JAMC-D-24-0154. Accepted.


Progress 02/15/23 to 02/14/24

Outputs
Target Audience:Scientific community (Meteorology, earth science, horticulturists, and Agricultural Engineering); Growers, crop consultants and farm/orchard managers. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Graduate students presented the technology specific results in local grower meetings (e.g., Washington State Tree Fruit Association (WSTFA) hosted Annual Meeting and NW Hort. Expo.), and at American Society of Agricultural and Biological Engineers AIM 2023 and were able to network with fellow researchers. Students also demonstrated the technology in Smart Orchard Field Day 2023 and to k-12 students and teachers during their visit to WSU Precision AG Lab. The PhD student translated research into teaching by presenting drone-sonde integration and frost mitigation use case as part of the department course BYSE 551 UAS in AG (2 credits). How have the results been disseminated to communities of interest?The results were shared within research community and grower stakeholders via presentations at ASABE AIM 2023, local grower meetings and as the media/news articles and interviews. The models to predict all critical weather variables for frost mitigation, results of drone-sonde, and components of localized intelligent crop loss management system with preliminary studies on radiative frost sensor were presented during local grower meetings. In 2022, AgWeatherNet conducted a Weather School where the recent findings of this experiment were shared. Videos of the presentations were edited in 2023, posted on AWN and CAHNRS you-tube channel, as well as shared through the following newsletters: WSU Viticulture, WSU Potatoes, Small Fruit Update. In 2024, the videos will be released monthly to the WSU tree fruit newsletter. These subscriptions collectively reach more than 3000 producers. The videos include the following: Climate: Why Does It Matter Meteorology Basics: Why They Matter AgWeatherNet Overview Weather Sensing: Principals & Tiers Soil Water Potential AgWeatherNet Web Portal Weather Sensing: Adding Private Stations to AWN (Hardware & Software) Weather Data Quality Assurance Cold Hardiness of Grapes, Blueberries and Cherries Private Weather Station Options (4 Companies) Forecasting Station-Specific Weather DAS - Potatoes DAS Overview What do you plan to do during the next reporting period to accomplish the goals? AIM# 1.Predictions of out-of-orchard reference station weather are useful, however, since weather conditions in the orchard can vary substantially from the cleared, flat, reference site, we have started quantifying both to understand how the reference station weather is related to in-orchard conditions. These efforts will continue in 2024 season. We plan to incorporate of Out-of-orchard to In-orchard weather prediction model onto the existing model framework to translate out-of-orchard forecasts to in-orchard forecasts using data collected in Spring 2023 and 2024. AIM# 2. To continuously monitor frost events, a Wireless Sensing Network (WSN) consisting of weather and radiative frost sensors will be distributed throughout the orchard block at cold spots, mapped by drone-sonde. The bud temperature distribution will be the primary step in any automated frost mitigation actuation system. The WSN will thus help drive decision support on activation of frost mitigation techniques. We have set up a semi-permanent, season-long observation site to collect 1.5 m (& 9 m) air temperature, relative humidity, and wind speed as well as bud temperatures and this will continue in Spring 2024. This will enable offset corrections to reference open field station conditions and estimate bud temperatures. The 1.5 and 9 m air temperatures will help estimate inversion strength in the orchards and validate such strengths from open field weather stations. Efforts will continue to evaluate the effectiveness of wind machines at varying distances (20m, 40m and 60m) and the impact on vertical air temperature structure to predict the effectiveness of wind machines under different temperature inversion scenarios. Drone- sondes and ground-based weather-sensing system will be used for this purpose. The efficacy of wind machines under different temperature inversion scenarios will also be investigated. The use of over and under tree sprinklers as a method for frost protection in Washington state is limited. Especially in highbush blueberries, growers have experienced branch breakage at the lower canopy sections due to excessive ice load and lingering moisture. Therefore, the suitability of automated sprinklers for frost protection in potted blueberry trees will be evaluated in spring 2024 (continuation from 2023 efforts. The effect of sprinkler droplet sizes in protecting the buds and in modifying the microclimate will be studied. Two micro-sprinkler emitter configurations suitable for highbush blueberry frost protection will be designed. The designed sprinkler system will be actuated based on air temperature and wind speed monitored using an infield weather sensor. The experiment will be conducted during frost nights and bud damage will be assessed in the laboratory, 24 hours after bud reaches room temperature. The designed emitter configurations will also be evaluated for spray coverage and deposition as a stand-in for pest management in summer 2024. AIM# 3. We will continue to engage with grower cooperators and share project knowledge at industry annual meetings and WSU-organized field days. If technology is ready for in-field demonstrations, we will conduct field days and survey the stakeholders to have meaningful feedback for path corrections (if any) to our research methods.

Impacts
What was accomplished under these goals? Aim# 1.Integrate surface and aerial meteorological observations into field-specific, short-term forecasts of weather variables relevant to frost mitigation. Continuing the efforts from 2022, to realize field-specific short-term weather forecasts, weather stations were deployed inside the orchard (3 sites) to collect year-long air temperature, relative humidity, wind speed, and inversion strength (1.5 and 9 m AGL) data. Such data is being used to re-train the site-specific weather forecasting models. Two small UAS integrated with a weather sensing hardware aka drone-sonde were used to simultaneously map vertical inversion strength in open field and in sweet cherry orchard block during frost nights. A retractable mast mounted on UTV was also deployed inside orchard to monitor vertical inversion strength during frost mitigation. Field campaigns were conducted on frost nights. Spatiotemporal temperature distribution mapping, using thermal infrared imagery, of three sweet cherry (cv. Chelan and cv. Bing) orchards (free-standing architecture) was also conducted. Results suggest that wind machines take 15 minutes (from starting point) to reach their maximum efficiency and the inversion strength degradation was inconsistent throughout the frost nights. The wind machines were able to increase the in orchard air temperature by about 1.9 °C at 20 m horizontally but were ineffective at 100 m away from their install location. The data collection will be continued for Spring 2024. Aim#2.Develop localized weather data-driven intelligent crop loss management system through real-time actuation of either- or combinations of active frost mitigation techniques, i.e., wind machines (frost fans), over-tree and under-tress fixed spray systems. A field deployable wireless network with sensing end-nodes capable of collecting wide-range of environmental sensor data was developed in sweet cherry orchard blocks. A small UAS comprising a thermal infrared imager was used to map the surface temperature of three sweet cherry orchard sites to identify cold spots, potential locations to monitor air and bud tissue temperature for frost mitigation. The spatiotemporal thermal maps were used to deploy sensing nodes encompassing air temperature, relative humidity probe and radiative frost sensor at cold spots for monitoring real-time air and bud tissue temperature using a wireless sensing network. At each site, temperature humidity probes were also deployed on respective wind machines. Results show that deployed wireless sensing network can extend up to 1000 m from centralized station with overall data loss of 9.7% with significantly low power. Statistically significant differences (5% level) were observed on air temperature recorded at wind machine compared to the cold spots. Also, data suggests that air temperature probs lag in quantifying actual bud temperature (by about 45 mins) and measuring bud tissue temperature, using radiative frost sensor or similar, is vital for frost management decision support. Trends signify the need to operational changes in current frost mitigation strategies. Validation studies will be conducted in Spring 2024. Pertinent to blueberry crop, a pilot experiment was conducted to on potted blueberries (c.v. Duke) to evaluate water based continuous, automated sprinkler, and automated micro sprinklers systems for frost protection. The experiment was conducted on a radiative frost night (2:00 am to 6:00 am) in 2023 season. There were 8.3% bud damage on continuous sprinkling and 16.4% for no mitigation. For Washington conditions, water freeze inside the irrigation pipes need to be a due consideration as growers explore these techniques. Team will continue to automate and test the conventional sprinkler and new micro-emitter based frost mitigation techniques (for blueberries) in 2024 field season. Aim#3. Assess grower evaluation/validation of decision aid tools and prototype performance. Our team has collected field data and tested the developed system prototypes on grower cooperator sites. Prior to the start of the experiments, the team met with early adopters of technologies and cooperators on cold hardiness projects. Their feedback in project objectives and design were incorporated in the experiments. Early spring field days are planned for the 2023-2024 frost season to showcase results and decision tools. Surveys and follow-up focus interviews will be conducted at the field day. Lastly, a presentation at the WA Small Fruit Conference on the use of cold hardiness models was followed by a circle group discussion with producers on ways to improve the interface and industry awareness of how to use decision tools better. The results of those discussions will be shared with the WSU AgWeatherNet team for improvements of decision tools.

Publications

  • Type: Other Status: Published Year Published: 2024 Citation: Gorthi, S. Hoheisel, G. A. and Khot*, L. R., 2024. Is it important to monitor bud tissue temperature for spring frost management in sweet cherry? Washington State University  Fruit Matters, February 1, 2024.
  • Type: Other Status: Published Year Published: 2023 Citation: Amogi, B. R., Schrader, M. J., Khot, L. R., Hoheisel, G. A., 2023. Decision Support Tools for Frost Mitigation. Washington State Tree Fruit Extension Fruit Matters, March 1, 2023. ~ 3000 subscribers. https://treefruit.wsu.edu/article/frost-mitigation/
  • Type: Other Status: Published Year Published: 2023 Citation: Amogi, B. R., Schrader, M. J., Khot, L. R., Hoheisel, G. A., 2023. Decision Support Tools for Frost Mitigation. Washington State Universitys electronic Viticulture and Enology Extension News (VEEN), Spring 2023 issue. https://s3-us-west-2.amazonaws.com/sites.cahnrs.wsu.edu/wp-content/uploads/sites/66/2023/04/17131707/2023-VEEN-Spring-FINAL.pdf
  • Type: Journal Articles Status: Submitted Year Published: 2023 Citation: Gorthi, S., L.R. Khot, and G. Hoheisel. 2023. Assessing direct and indirect methods to monitor bud tissue temperature for effective spring frost management in Sweet Cherry (Prunus avium cv. Chelan). Agricultural and Forest Meteorology. (Submitted).
  • Type: Conference Papers and Presentations Status: Published Year Published: 2023 Citation: Gorthi, S., B.R. Amogi, G. Hoheisel, and L.R. Khot. 2023. Wireless sensing network for frost mitigation in sweet cherry orchards. Paper No. 2300549, ASABE 2023 Annual International Meeting, Omaha, NE, July 9-12, 2023 (Oral Presentation). Time: 5 min. Participants ~100.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2023 Citation: Gorthi, S., K. Yumnam, B. R. Amogi, G. Hoheisel, and L.R. Khot. 2023. Evaluation of contact/non-contact type bud temperature monitoring sensors for frost management in sweet cherry crop. Paper No. 2300551, ASABE 2023 Annual International Meeting, Omaha, NE, July 9-12, 2023 (Oral Presentation). Time: 15 min. Participants ~100.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2023 Citation: Yumnam, K, S. Gorthi, B.R. Amogi, G. Hoheisel, and L.R. Khot, 2023. Drone-sonde and ground sensing-based evaluation of wind machines effectiveness in sweet cherry spring frost mitigation. Paper No. :2301395, ASABE 2023 Annual International Meeting, Omaha, NE, July 9-12, 2023 (Oral Presentation). Time: 15 min. Participants ~100.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2023 Citation: Schrader, M. J., B. R. Amogi, and L.R. Khot, 2023. Dronesonde based temperature inversion strength monitoring and grower decision support tool for frost mitigation in perennial specialty crops. Paper No. 2301273, ASABE 2023 Annual International Meeting, Omaha, NE, July 9-12, 2023 (Oral Presentation).
  • Type: Conference Papers and Presentations Status: Published Year Published: 2023 Citation: Gorthi, S., L.R. Khot, G. Hoheisel. 2023. Assessment of Internet of Things based frost monitoring network in sweet cherry orchard. Presented at the Annual Meeting of the Washington State Tree Fruit Association, Kennewick, WA. December 46, 2023 (Oral Presentation). Time: 5 min. Participants ~150.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2023 Citation: Schrader*, M.J., Gorthi*, S., L. R. Khot 2023. Frost Decision Aid Tools in Permanent Cropping Systems. Smart Orchard Field Day. Wenatchee, WA. September 15, 2023. Time: 10 min. Participants: ~40.


Progress 02/15/22 to 02/14/23

Outputs
Target Audience:Scientific community (Meteorology, earth science, horticulturists and Agricultural Engineering); Growers, crop consultants and farm managers. Changes/Problems:Postdoc Dr. Matt Cann left WSU in December 2022 for industry job which has caused some delay in activities until a new postdoc could be hired. WSU HR is helping us with the recruitment. What opportunities for training and professional development has the project provided?Dr. Matt Cann (postdoc) has presented and networked at two conferences in the past year. Dr. Cann has obtained mentoring experience in field experiment planning and data analysis with PhD students. The PhD students worked with grower cooperators as they deployed sensing systems for field-data collection, experimentation, thus understanding real-orchard decision support scenario's and futuristic technology needs. Additionally, PhD students presented the results in local grower meetings (e.g., Washington State Tree Fruit Association (WSTFA) hosted 118th Annual Meeting and NW Hort Expo) and at American Society of Agricultural and Biological Engineers AIM 2022 and were able to network with fellow researchers. One of the PhD students also attended annual PI meeting, presented project poster and interacted with fellow researchers. The PhD student translated research into teaching by presenting drone-sonde integration and frost mitigation use case as part of the department course BYSE 552 UAS in AG (2 credits). How have the results been disseminated to communities of interest?The results were shared within research community and grower stakeholders via presentations at the American Meteorological Society at conference, the AgWeatherNet Weather School 2022, ASABE AIM 2022, and as the media/news articles and interviews. The models to predict all critical weather variables for frost mitigation, results of drone-sonde, and components of localized intelligent crop loss management system with preliminary studies on radiative frost sensor were presented at the 2022 CPS PI Meeting. What do you plan to do during the next reporting period to accomplish the goals?AIM# 1. Predictions of out-of-orchard reference station weather are useful, however, since weather conditions in the orchard can vary substantially from the cleared, flat, reference site, we must measure both to understand how the reference station weather is related to in-orchard conditions. Therefore, this year, we plan to incorporate of Out-of-orchard to In-orchard weather prediction model onto existing model framework to translate out-of-orchard forecasts to in-orchard forecasts using data collected in Spring 2023. AIM# 2. A Wireless Sensing Network (WSN) consisting of weather and radiative frost sensors will be distributed throughout the orchard block at cold spots, mapped by drone-sonde, to continuously monitor frost events. The bud temperature distribution will be the primary step in any automated frost mitigation actuation system. The WSN will thus help drive decision support on activation of frost mitigation techniques. We plan to set up a semi-permanent, season-long observation sites to collect 1.5 m ( & 9 m) air temperature, relative humidity, and wind speed as well as bud temperatures in Spring of 2023. This will enable offset corrections to reference open field station conditions and estimate bud temperatures. The 1.5 and 9 m air temperature will help estimate inversion strength in the orchards and validation of such with strengths from open field weather stations. We plan to evaluate effectiveness of wind machines at varying distances (20m, 40m and 60m) and impact on vertical air temperature structure to predict effectiveness of wind machines under different temperature inversion scenarios. Drone-sondes and ground-based based weather sensing system will be used for this purpose. The efficacy of wind machines under different temperature inversion scenarios will also be investigated. The use of over and under tree sprinklers as a method for frost protection in Washington State is limited. Especially in highbush blueberries, growers have experienced branch breakages due to excessive ice load and moisture lingering at the lower canopy sections. Therefore, the suitability of automated sprinklers for frost protection in blueberry fields will be evaluated in 2023. The effect of sprinkler droplet sizes in protecting the buds and in modifying the microclimate will be studied. AIM# 3. We will continue to engage with grower cooperators and share project knowledge in the annual meetings/WSU organized field days. If technology is ready for in-field demonstrations, we will conduct field days and survey the stakeholders to have meaningful feedback for path corrections (if any) to our research methods.

Impacts
What was accomplished under these goals? Aim# 1. Integrate surface and aerial meteorological observations into field-specific, short-term forecasts of weather variables relevant to frost mitigation. A machine-learning based next-day weather forecast postprocessing model was designed, built, and tested operationally for AgWeatherNet stations to predict weather variables relevant to frost mitigation. Results from real-time testing show the postprocessing model is comparable or superior to third-party weather forecast providers when predicting next-day low temperatures, the most important variable for frost mitigation. The next-day minimum temperature model was expanded into an hourly 10-day model with full suite of weather variables, and an hourly updated 36-hour nowcast model. These models predict all critical weather variables for frost mitigation including air temperature, inversion strength, and wind speed. A small UAS aka drone-sonde has been integrated with a weather sensing hardware and a thermal infrared imager to map the vertical inversion strengths as well as spatiotemporal temperature distribution. The drone-sonde was used for in-field mapping of the vertical inversion strengths in 2022 field season. Field campaigns were conducted on frost nights. Preliminary observations suggest that mapping during drone-sonde ascent were more consistent than descent. Spatiotemporal temperature distribution mapping, using thermal infrared aerial imagery, of two sweet cherry (cv. Bing and cv. Chelan) orchards (free-standing architecture); one with 12% slope and the other with 5% slope) was also conducted. Data suggest that combined effects of wind machine and under-tree sprinkler operations managed to limit the temperature above the critical damage temperature. Additionally, for orchards located on 12% slopes, the early morning orchard surface temperature maps along with in-orchard weather data quantified spatiotemporal variations. Trends signify the need to operational changes in current frost mitigation strategies. The data collection will be continued in Spring 2023. Aim#2. Develop localized weather data-driven intelligent crop loss management system through real-time actuation of either-or combinations of active frost mitigation techniques, i.e. wind machines (frost fans), over-tree and under-tress fixed spray systems. A field deployable prototype sensing node comprised of thermal-RGB imaging, radiation frost sensor (leaf and bud temperature senor) and all-in-one weather sensor has been developed. These nodes were deployed in Sweet Cherry orchards (3 sites) in 2022 field season to quantify data suitability of the type sensor to be integrated with wireless sensor network to have real-time bud temperature monitoring and forecasting (in future). collection in 2022 season . Overall, data analysis suggest that a non-contact type bud temperature sensor can accurately mimic and estimate actual bud temperature. The difference between air and actual bud temperature was 0.2 to 3 °C. However, the average difference between the bud temperature sensor and the actual bud temperature was less than 0.5 °C. Therefore, the decision-support system on sensing nodes shall include bud temperature sensor measurements as a constraint instead of air temperature. A high correlation between non-contact and contact type bud temperature sensor was observed. During the field experiments we found that measuring internal bud temperature using contact type sensor was challenging. Therefore, we recommend integrating radiative frost sensor into decision support in addition to air temperature and relative humidity. Aim#3. Assess grower evaluation/validation of decision aid tools and prototype performance. Our team is collected field data and test the developed system prototypes on grower cooperator sites. More robust evaluation/validation of decision aid tools and prototypes will be conducted in 2023 and 2024 field season.

Publications

  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Amogi, B. R., M. Cann, G. Kothawade, S. Gorthi, K. Yumnam, G.-A. Hoheisel, Lav R. Khot. 2022. Paper No. 2201213, ASABE 2022 Annual International Meeting, Houston, TX, July 17-20, 2022 (Oral Presentation).
  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Cann, M. D., J. P. Boomgard-Zagrodnik, and D. Brown (2022, January 25). Site-Specific ML Postprocessing of HRRR Output for Short-Term Frost Forecasts and Nowcasts in Complex Terrain [Conference Presentation]. American Meteorological Society 2022 Annual Meeting, Houston, TX, United States. https://ams.confex.com/ams/102ANNUAL/meetingapp.cgi/Paper/398682
  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Cann, M. D., (2022, February 4). Forecasting Station-Specific Weather (Heat and Cold Stress) [Conference Presentation]. Washington State University Weather School 2022, Pullman, WA, United States.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Cann, M. D., B. R. Amogi, S. Gorthi, G. A. Hoheisel, and L. R. Khot, (2022, June 27) Observing and Forecasting Near-Surface Temperature Inversions for Effective frost Mitigation in Central WA Perennial Specialty Crops [Conference Presentation]. American Meteorological Society 20th Conference on Mountain Meteorology, Park City, UT, United States. https://ams.confex.com/ams/20MOUNTAIN/meetingapp.cgi/Paper/402770
  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Gorthi* S., K. Yumnam, B. R. Amogi, M.T. Cann, L.R. Khot, and G. Hoheisel. 2022. Evaluation of various sensors for frost management in Sweet Cherries. 118th Annual Meeting & NW Hort. Expo., Wenatchee, WA. December 68, 2022 (Oral Presentation). Time: 5 min. Participants ~150.


Progress 02/15/21 to 02/14/22

Outputs
Target Audience:Scientific community (Meteorology, earth science, horticulturists and Agricultural Engineering); Growers, crop consultants and farm managers Changes/Problems:The project needed PI change, as Dr. Dave Brown and Dr. Joe Zagrodnikmoved to the private industry.Updated PIs list is as below Lav R. Khot, Gwen-Alyn Hohseisel. Team also faced challenges hiring graduate students due to COVID-19 scenarios. What opportunities for training and professional development has the project provided?Post-doctorate researcher Matthew Cann has begun gaining experience working with students in a supervisory/mentor role. Matthew Cann also presented relevant work to the scientific community at the American Meteorological Society Annual Meeting in January 2022 and to AgWeatherNet stakeholders at the WSU (Washington State University) Weather School in February 2022. Graduate student was provided opportunity to present the weather mapping technology to the international collaborators visiting WSU as part of California-Washington-Dutch collaboration. How have the results been disseminated to communities of interest?Presentations of relevant and future work were given to a broad range of weather scientists at the American Meteorological Society Annual Meeting in January and to the grower stakeholders at the WSU Weather School 2022. The work was also shared with horticulture?and agricultural engineering community via invited talks duringAnnual Meeting of American Society of Horticultural Sciences, Denver, CO, USA andSpecialty Crop Engineering: Advanced Technologies for Specialty Crop Production, ASABE Annual International Meeting(Virtual). What do you plan to do during the next reporting period to accomplish the goals?The machine-learning model forecasts of station-specific weather variables relevant to frost mitigation will be transformed to predict weather at finer timescales with forecasts produced multiple times per day. The station-specific weather forecasts will then be interpolated to specific sites in the field where weather stations will be added to improve model accuracy over time. Additionally, an array of bud temperature sensors and temperature inversion measurements will be made to provide robust statistical relationships to the installed weather station and weather forecasts. In 2022 field season, we will study wind machine effect on temperature profiles with drones during spring and fall frost/freeze events. We will also deploy the prototype in-field weather and crop physiology sensing nodes in sweet cherry and blueberries fields to develop understanding on real-time sensing of bud temperature and inversion strengths for appropriate actuation of frost mitigation (e.g., using wind machines). We will continue documenting current frost management practices and associated inefficiencies.

Impacts
What was accomplished under these goals? Aim# 1. A machine-learning based next-day weather forecast postprocessing model was designed, built, and tested operationally for AgWeatherNet stations to predict weather variables relevant to frost mitigation. Results from real-time testing show the postprocessing model is comparable or superior to third-party weather forecast providers when predicting next-day low temperatures, the most important variable for frost mitigation. A small UAS has been integrated with a weather sensing hardware and a thermal infrared imager to map the vertical inversion strengths as well as spatiotemporal temperature distribution. This system has been pilot tested and will be used for in-field mapping in 2022 field season. Aim#2. A field deployable prototype sensing node comprised of thermal-RGB imaging, radiation frost sensor (leaf and bud temperature senor) and all-in-one weather sensor has been developed. The integrated system will be used for in-field data collection in 2022 season. Aim#3.

Publications

  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Cann, M. D., J. P. Boomgard-Zagrodnik, and D. Brown (2022, January 25). Site-Specific ML Postprocessing of HRRR Output for Short-Term Frost Forecasts and Nowcasts in Complex Terrain [Conference Presentation]. American Meteorological Society 2022 Annual Meeting, Houston, TX, United States. https://ams.confex.com/ams/102ANNUAL/meetingapp.cgi/Paper/398682
  • Type: Other Status: Published Year Published: 2022 Citation: Cann, M. D., (2022, February 4). Forecasting Station-Specific Weather (Heat and Cold Stress) [Conference Presentation]. Washington State University Weather School 2022, Pullman, WA, United States.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Khot, L.R. 2021. (Invited) Workshop Presentation on Sensing (in-field/remote) Technologies for Abiotic Crop Stress Monitoring, Environmental instrumentation in orchard/vineyard management, Annual Meeting of American Society of Horticultural Sciences, Denver, CO, USA. August 9, 2021.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Khot, L.R. 2021.(Invited) Session Talk Precision Ag & Automation Technologies for Specialty Crop Production Management, Specialty Crop Engineering: Advanced Technologies for Specialty Crop Production, ASABE Annual International Meeting, July 12, 2021. (Virtual).
  • Type: Other Status: Published Year Published: 2021 Citation: Khot, L.R. 2021. (Invited) Seminar on Precision Ag & Automation Technologies for Specialty Crop Production Management, Summerland Research and Development Centre, Agriculture and Agri-Food Canada; March 30, 2021.