Progress 06/01/22 to 05/31/23
Outputs
Target Audience:Growers in southern California desert agriculture region Scientific community Cooperative extension personnel Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?One PhD studentwas trained and has been involved in this project. How have the results been disseminated to communities of interest?We delivered a presentation in a UCANR CE workshop organized by the co-PI of this project for the growers in the region, focusing on the IMT-Desert decision support tool and how to use it for irrigation and salinity management of major crops. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
IMT-DESERT decision support tools were developed for the growers in the region. IMT-DESERT: AN IRRIGATION MANAGEMENT TOOL FOR SOUTHERN CALIFORNIA DESERT REGION The IMT-Desert is a web-based application based on crop water production functions for 12 major crops in the region. The tool also estimates leaching requirements based on soil type, crop type, and salinity information using standard and SALEACH models. In the advanced mode, the IMT-Desert can compare the yield and revenue of up to five crops simultaneously based on the user input about irrigation water salinity, available water quantity, and the irrigation system's efficiency. In addition: Multiple easy-to-use remote-sensing-based regression models were developed to estimate alfalfa yield via multispectral and thermal drone imagers. Predicting crop yield and generating yield maps allow growers to determine the within and between fields yield variabilities and apply site-specific variable rates of agricultural inputs as needed. Currently, mainstream growers in southern California do not have access to yield monitor systems that produce yield maps, a crucial piece of information to move toward precision irrigation management. Yield monitor systems are either unavailable for the crops grown in the region or are not widely adopted by the growers. An alternative approach for yield estimation is remote sensing via drones, which is gaining popularity because of its ease of use and potential for rapid measurements from the field to the regional scale. Our results showed promising performance for regression-drone-based models to estimate alfalfa yield and differentiate among fields with low and high productivity in the region. Goal II: Our HYDRUS modeling results showed the performance of traditional uniform irrigation depends on the field-level soil texture variability and the area of each soil-texture-based management zone. Variable rate irrigation (VRI) scenario can significantly decrease water loss through deep percolation when one of the major soil textures in the field has a limited plant-available water-holding capacity (e.g., loamy sand). However, when the primary soil textures had enough plant-available water holding capacity, the performance of uniform irrigation was comparable to VRI in terms of higher transpiration and lower water loss. Goal III: Already completed Goal IV: Presentation for growers in the region
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2023
Citation:
Sapkota, A*., Haghverdi, A*., Montazar, A. (2023). Estimating fall-harvested alfalfa (Medicago sativa L.) yield using UAV-based multispectral and thermal images in southern California. Agrosystems, Geosciences & Environment .2023;6:e20392.
|
Progress 06/01/19 to 05/31/23
Outputs
Target Audience:Growers in southern California desert agriculture region Scientific community Cooperative extension personnel Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Two PhD students, one postdoctoral scholar and one undergraduate student researcher were trained and have been involved in this project. How have the results been disseminated to communities of interest?Results from the project thus far have been disseminated to communities mainly through presentations at scientific conferences and through extension presentations to growers. We also used the group website and Twitter account to post recorded presentations and upload educational materials related to this project. We delivered five presentations in three UCANR CE workshops that wereco-organized with the co-PI of this project for the growers in the region. Weprovided the workshop participantswith updates regarding the results of the project and how to use the decision-support tools we developed. In addition, we worked with multiple alfalfa growers in the region to collect data to develop easy-to-use statistical models for yield estimation using drone-based multispectral images. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Two decision support tools were developed for the growers in theregion. Goal I: (I) SAMZ-DESERT: A SATELLITE-BASED AGRICULTURAL MANAGEMENT ZONES TOOL FOR SOUTHERN CALIFORNIA DESERT REGION The SAMZ-Desert is a GIS-based tool that uses Landsat 8 NDVI data to provide information about within-field heterogeneity and how to address it by delineating management zones for nearly 7,000 agricultural fields. (II) IMT-DESERT: AN IRRIGATION MANAGEMENT TOOL FOR SOUTHERN CALIFORNIA DESERT REGION The IMT-Desert is a web-based application based on crop water production functions for 12 major crops in the region. The tool also estimates leaching requirements based on soil type, crop type, and salinity information using standard and SALEACH models. In the advanced mode, the IMT-Desert can compare the yield and revenue of up to five crops simultaneously based on the user input about irrigation water salinity, available water quantity, and the irrigation system's efficiency. In addition: Multiple easy-to-use remote-sensing-based regression models were developed to estimate alfalfa yield via multispectral and thermal drone imagers.Predicting crop yield and generating yield maps allow growers to determine the within and between fields yield variabilities and apply site-specific variable rates of agricultural inputs as needed. Currently, mainstream growers in southern California do not have access to yield monitor systems that produce yield maps, a crucial piece of information to move toward precision irrigation management. Yield monitor systems are either unavailable for the crops grown in the region or are not widely adopted by the growers. An alternative approach for yield estimation is remote sensing via drones, which is gaining popularity because of its ease of use and potential for rapid measurements from the field to the regional scale. Our results showed promising performance for regression-drone-based models to estimate alfalfa yield and differentiate among fields with low and high productivity in the region. Goal II: Our HYDRUS modeling results showed the performance of traditional uniform irrigation depends on the field-level soil texture variability and the area of each soil-texture-based management zone. Variable rate irrigation (VRI) scenario can significantly decrease water loss through deep percolation when one of the major soil textures in the field has a limited plant-available water-holding capacity (e.g., loamy sand). However, when the primary soil textures had enough plant-available water holding capacity, the performance of uniform irrigation was comparable to VRI in terms of higher transpiration and lower water loss. Goal III: We conducted a regional statistical analysis on more than 6000 fields in the region. A total number of 11 cloud-free images in 2018-2020 were statistically analyzed to determine the extent of within-field NDVI variability and temporal stability of MZs at the regional level. A majority (approx. 37%) of the fields had four zones as an optimum number of zones in the region, which could address more than 85% of the within-field NDVI variability. Around 13% (n = 873) of the fields in the region were strongly spatially-clustered in at least half the Landsat-8 cloud-free scenes and can benefit from variable rate technologies. Our results suggest that dynamic zoning over time might be necessary for most of these fields. Goal IV: We organized workshops and delivered presentationsfor the growers in the region to disseminate the Results of the work. In addition, we worked with multiple alfalfa growers in the region to collect data to develop easy-to-use statistical models for yield estimation using drone-based multispectral images.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2023
Citation:
Sapkota, A*., Haghverdi, A*., Montazar, A. (2023). Estimating fall-harvested alfalfa (Medicago sativa L.) yield using UAV-based multispectral and thermal images in southern California. Agrosystems, Geosciences & Environment .2023;6:e20392.
- Type:
Journal Articles
Status:
Published
Year Published:
2022
Citation:
Garg, A., Sapkota, A., Haghverdi, A*. (2022). SAMZ-Desert: A Satellite-based Agricultural Management Zoning Tool for the Desert Agriculture Region of Southern California. Computers and Electronics in Agriculture. 194 (2022) 106803.
|
Progress 06/01/21 to 05/31/22
Outputs
Target Audience:Southern California desert growers, cooperative extension personnel, the scientific community Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Two PhD students, one postdoctoral scholar and one undergraduate student researcher were trained and have been involved in this project. How have the results been disseminated to communities of interest?Results from the project thus far have been disseminated to communities mainly through presentations in the scientific conferences and through extension presentations to the growers. We also used the group website and Twitter account to post recorded presentations and upload educational materials related to this project. We delivered a presentation in a UCANR CE workshop organized by the co-PI of this project for the growers in the region to provide them with updates regarding the results of the project.In addition, we worked with multiple alfalfa growers in the region to collect data to develop easy-to-use statistical models for yield estimation using drone-based multispectral images. What do you plan to do during the next reporting period to accomplish the goals?We will finish the HYDRUS modeling task and finish and publish the second decision support tool. We also submit multiple extension publications based on the results of this work. We will also submit at least two more technical journal articles based on the results of this project.
Impacts
What was accomplished under these goals?
Goal-1: (a) The UCRWATER team developed a satellite-based agriculture management zoning tool for the southern California desert agriculture region, called "SAMZ-Desert". This is a web-GIS-based interactive tool that shows Landsat-NDVI-based MZs for all the fields in southern California. This tool was developed for end-users and growers with no experience and expertise required to obtain the results for individual fields. The users only need to find their fields (using a google map embedded in the tool), click on the map and see the results. All the complicated remote sensing and statistical analysis steps happen behind the scene. We have been also working on a second toolto help growers in identifying optimum irrigation management strategies to maximize the crop selection,yield and irrigation water use efficiency.SAMZ-Desert can be accessed from the Haghverdi Water Management Group website: http://www.ucrwater.com/software-and-tools.html. Goal-2: Significant progress was ade toward HYDRUS modeling to understand the strength and benefits of site-specific irrigation scenarios in the region. Goal-3: We conducted a regional statistical analysis on more than 6000 fields in the region.A total number of 11 cloud-free images in 2018-2020 were statistically analyzed to determine the extent of within-field NDVI variability and temporal stability of MZs at the regional level. A majority (approx. 37%) of the fields had four zones as an optimum number of zones in the region, which could address more than 85% of the within-field NDVI variability. Around 13% (n = 873) of the fields in the region were strongly spatially-clustered in at least half the Landsat-8 cloud-free scenes and can benefit from variable rate technologies. Our results suggest that dynamic zoning over time might be necessary for most of these fields. Goal-4: We delivered a presentation in a workshop organized by the co-PI for the growers in the region to disseminate the results of the work. In addition, we worked with multiple alfalfa growers in the region to collect data to develop easy-to-use statistical models for yield estimation using drone-based multispectral images.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2022
Citation:
Garg, A., Sapkota, A., Haghverdi, A*. (2022). SAMZ-Desert: A Satellite-based Agricultural Management Zoning Tool for the Desert Agriculture Region of Southern California. Computers and Electronics in Agriculture. 194 (2022) 106803.
|
Progress 06/01/20 to 05/31/21
Outputs
Target Audience:Southern California desert growers, cooperative extension personnel, the scientific community Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Two PhD students,one postdoctoral scholar and one undergraduate student researcher were trained and have been involved in this project. How have the results been disseminated to communities of interest?Results from the project thus far have been disseminated to communities mainly through presentations in the scientific conferences and through extension presentations to the growers. We also used the group website and Twitter account to post recorded presentations and upload educational materials related to this project. What do you plan to do during the next reporting period to accomplish the goals?During the last year of the project, we will draft and submit multiple outreachpapers and technical journal articles. We will finish field data collection and laboratory analysis. We will finishthe development of the decision support tools and publish them. We also plan toorganize at least one more workshop for the growers in the region. We also finish the regional assessment analysis and HYDRUS modeling tasks.
Impacts
What was accomplished under these goals?
Goal-1: (a) A decision support tool is in the process of development to help growers in identifying optimum management strategies to maximize the crop yield. Multiple crops have been included in the tool to quantify their yield response to water deficits. The tool also considers the crop growth stages in order to be able to direct more water to a particular stage as compared to others based on the water requirements or crop sensitivity under deficit irrigation conditions. (b) A decision support tool (https://arcg.is/144zG11) has been developed. This helps to identify if the field needs to be delineated into management zones or not for site-specific irrigation. The decision support tool was based on the satellite images, ground truth data are recently collected to quantify the accuracy of using satellite images. This tool will be updated as needed. Goal-2: A new PhD student joined the research group and started simulation using the HYDRUS model. Goal-3: To be completed this year. Goal-4: We organized a virtual workshop for growers in the region in collaboration with the co-PI Dr. Montazar and deliverd the following presentations: Sapkota, A. and A. Haghverdi. Challenges and opportunities of using drones for irrigation management. Webinar: Irrigation management tools and technologies. University of California Cooperative Extension. March 2021 (approx. 40 attendees) Garg, A., A. Sapkota, and A. Haghverdi. VRI-EVAL: A web-based tool for variable rate irrigation pre-adoption assessment. Webinar: Irrigation management tools and technologies. University of California Cooperative Extension. March 2021 (approx. 40 attendees) Haghverdi, A. Fundamentals of site-specific variable rate irrigation management. Webinar: Irrigation management tools and technologies. University of California Cooperative Extension. March 2021 (approx. 40 attendees)
Publications
|
Progress 06/01/19 to 05/31/20
Outputs
Target Audience:
Nothing Reported
Changes/Problems:The spread of Covid-19 and UCR campus closure slowed down our research and forced the team to postpone on-farm data collection for several months. What opportunities for training and professional development has the project provided?~ One PhD student and one postdoctoral scholar were recruited and have been trained and working on the project. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?~ We will develop the first prototype of the zone delineation decision support tool. ~ We will draft extension publication on fundamentals of site-specific variable rate irrigation management. ~ Finish the literature review and start prototyping the water production function decision support tool. ~More intense data collection from growers fields
Impacts
What was accomplished under these goals?
Impact of the project: Irrigation drives southern California desert agricultural systems. Imperial County, for example, produced $20 billion gross in 2017 with more than half a million acres of crop production (Imperial County Agricultural Crop and Livestock Report, 2017). On-farm water conservation programs will continue to be the major source of water to meet the transfer obligations, making new irrigation management technologies all the more critical. We primarily focused on objectives 1 and 3 during the reporting period. Task 1: Development of irrigation water optimizer tool (IWO): a. Major activities completed/experiments conducted An extensive literature search was done using online databases to find peer-reviewed papers that studied the effect of irrigation in crop production in the Southern California region and have developed the relationship between irrigation and crop production. Few studies were identified which were done in the region and have reported the relationship between the water application and crop production. b. Data collected Based on the literature search, available water production functions of various crops in Southern California were identified. The irrigation-production equations of the following crops were collected - sunflower, oats, tomato, cauliflower, wheat, sorghum, lettuce, barley, cowpea, and celery. Task 2: Development of VRI Pre-Adoption Assessment Tool (VRI-EVAL) a. Major activities completed/experiments conducted A prototype for GIS-based decision support tool (Variable Rate Irrigation Pre-Adoption Assessment Tool, VRI-EVAL) has been developed to perform pre-adoption assessment for emerging site-specific variable rate irrigation (VRI) systems. VRI-EVAL will help growers in optimally dividing the fields into homogeneous subregions (irrigation management zones (MZs)) based on irrigation requirements to identify optimum irrigation management strategies under uniform full/deficit irrigation conditions. For this purpose, multiple fields in southern California with substantial variability are selected as study areas to conduct experiments. Several satellite-based crop indices, including, normalized difference vegetation index (NDVI), reference evapotranspiration (ETrf), and digital elevation model (DEM) are utilized to delineate the MZs of the fields. The VRI-EVAL is designed with some statistical filters to help users check the quality of farming data and remove outliers and bad data prior to using the data for irrigation management zone delineation. The Statistical multivariate clustering techniques (like, k-means) are then applied on the filtered input using Python and ArcGIS Pro. Then the performance of the zoning approach is evaluated based on the inter-zone variance approach to explain percent of variability within zones and determine the optimum number of irrigation MZs. After that, the irrigated fields are spatially arranged into an optimal number of MZs using majority filtering technique based on the irrigation strategies used. The results obtained using satellite data will be validated with the help of drone data collected at fields. The VRI-EVAL is created by adding elements which use all the information and statistics computed for the study area for interactive visualization and providing vital insights. b. Data collected Cloud free LANDSAT satellite images for the year 2019 were collected from USGS EarthExplorer (https://earthexplorer.usgs.gov/) and METRIC EEFLUX (https://eeflux-level1.appspot.com/) to conduct experiments on the area of interest (fields of southern California). Multispectral and thermal images were also collected using drones from the producer field at Imperial County, CA. The desktop GIS application from Esri (ArcGIS Pro), and python programming language are used to process the satellite data. Similarly, the drone images are processed in Pix4D to create a desired orthomosaic of the study site. c. Summary statistics and discussion of results Preliminary analysis carried out at multiple irrigated fields in southern California using satellite images showed that the tool (VRI-EVAL) provides satisfactory performance in terms of visualization and providing key insights for at-a-glance decision making, and monitor activities. VRI-EVAL is fast, interactive, easily adaptable, user-friendly, and freely available. It displays Google Earth imagery with highlighted areas/fields of interest, along with various elements added, such as, optimal number of zones indicator, gauge showing statistics of selected input satellite product, images of fields arranged into optimal number of zones and performance graph for a selected field. In addition, it gives warning messages, such as "zoning is not required", if the field is almost uniform. Most of the elements of the tool are interactive to each other. The tool will also provide flexibility to the users/farmers to select the feasible number of irrigation MZs by having a look at different available options. The tool works for both circular and rectangular fields representing a wide range of irrigation systems, including center pivot, lateral move, surface, and drip irrigation systems. Outcomes will be validated using the drone images collected and then scaled up to all the fields in Southern California with many more elements and flexibility added to VRI-EVAL. Task 4: Determine the potential adoption of VRI systems in the southern California desert agricultural region through an extensive geospatial analysis a. Major activities completed/experiments conducted To evaluate the potential adoption of variable rate irrigation systems (VRI) in southern California, first, we randomly selected a few center pivot irrigated fields in the region. We then used satellite products with different indices including, normalized difference vegetation index (NDVI), reference evapotranspiration (ETrf), and digital elevation model to delineate the management zones (MZ) of the field. The delineation of MZs was performed using multivariate clustering (k-means) in ArcGIS Pro. After this preliminary evaluation, we selected three different producer fields in Imperial county with the help from the co-PI Dr. Ali Montazar. These fields are planted with forage and vegetable crops. From one of the producer's fields, we collected drone images. We flew DJI Matrice-100 mounted with a multispectral (Micasense RedEdge) and a thermal (FLIR Duo Pro R) camera. Images from this preliminary flight have been cleaned and processed for further analysis. We will use these images, and images collected in the future to validate the results obtained after processing satellite data. b. Data collected Cloud free LANDSAT satellite images for the year 2019 and 2020 were collected from USGS EarthExplorer (https://earthexplorer.usgs.gov/) and METRIC EEFLUX (https://eeflux-level1.appspot.com/) to perform geospatial analysis of the area of interest. We also collected multispectral and thermal images using drones from the producer field at Imperial County, CA. We processed satellite images in ArcGIS Pro using an unsupervised clustering algorithm. The efficacy of site-specific VRI was evaluated using a variance-based evaluation approach. Similarly, the drone images are processed in Pix4D to create a desired orthomosaic of the study site. c. Summary statistics and discussion of results Preliminary analysis done at multiple irrigated fields in southern California using satellite images showed that dividing a field into three to four management zones can address up to 90% of the within-field variability. This shows the significant potential of VRI in the region. Findings will be validated using images collected by drones and then scaled up to all the fields in Southern California. ?A screenshot of the prototype of VRI-EVAL tool is shown below for reference purpose:
Publications
|
Progress 04/01/19 to 03/31/20
Outputs
Target Audience:
Nothing Reported
Changes/Problems:The spread of Covid-19 and UCR campus closure slowed down our research and forced the team to postpone on-farm data collection for several months. What opportunities for training and professional development has the project provided?~ One PhD student and one postdoctoral scholar were recruited and have been trained and working on the project. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?~ We will develop the first prototype of the zone delineation decision support tool. ~ We will draft extension publication on fundamentals of site-specific variable rate irrigation management. ~ Finish the literature review and start prototyping the water production function decision support tool. ~More intense data collection from growers fields
Impacts
What was accomplished under these goals?
Impact of the project: Irrigation drives southern California desert agricultural systems. Imperial County, for example, produced $20 billion gross in 2017 with more than half a million acres of crop production (Imperial County Agricultural Crop and Livestock Report, 2017). On-farm water conservation programs will continue to be the major source of water to meet the transfer obligations, making new irrigation management technologies all the more critical. We primarily focused on objectives 1 and 3 during the reporting period. Task 1: Development of irrigation water optimizer tool (IWO): a. Major activities completed/experiments conducted An extensive literature search was done using online databases to find peer-reviewed papers that studied the effect of irrigation in crop production in the Southern California region and have developed the relationship between irrigation and crop production. Few studies were identified which were done in the region and have reported the relationship between the water application and crop production. b. Data collected Based on the literature search, available water production functions of various crops in Southern California were identified. The irrigation-production equations of the following crops were collected - sunflower, oats, tomato, cauliflower, wheat, sorghum, lettuce, barley, cowpea, and celery. Task 2: Development of VRI Pre-Adoption Assessment Tool (VRI-EVAL) a. Major activities completed/experiments conducted A prototype for GIS-based decision support tool (Variable Rate Irrigation Pre-Adoption Assessment Tool, VRI-EVAL) has been developed to perform pre-adoption assessment for emerging site-specific variable rate irrigation (VRI) systems. VRI-EVAL will help growers in optimally dividing the fields into homogeneous subregions (irrigation management zones (MZs)) based on irrigation requirements to identify optimum irrigation management strategies under uniform full/deficit irrigation conditions. For this purpose, multiple fields in southern California with substantial variability are selected as study areas to conduct experiments. Several satellite-based crop indices, including, normalized difference vegetation index (NDVI), reference evapotranspiration (ETrf), and digital elevation model (DEM) are utilized to delineate the MZs of the fields. The VRI-EVAL is designed with some statistical filters to help users check the quality of farming data and remove outliers and bad data prior to using the data for irrigation management zone delineation. The Statistical multivariate clustering techniques (like, k-means) are then applied on the filtered input using Python and ArcGIS Pro. Then the performance of the zoning approach is evaluated based on the inter-zone variance approach to explain percent of variability within zones and determine the optimum number of irrigation MZs. After that, the irrigated fields are spatially arranged into an optimal number of MZs using majority filtering technique based on the irrigation strategies used. The results obtained using satellite data will be validated with the help of drone data collected at fields. The VRI-EVAL is created by adding elements which use all the information and statistics computed for the study area for interactive visualization and providing vital insights. b. Data collected Cloud free LANDSAT satellite images for the year 2019 were collected from USGS EarthExplorer (https://earthexplorer.usgs.gov/) and METRIC EEFLUX (https://eeflux-level1.appspot.com/) to conduct experiments on the area of interest (fields of southern California). Multispectral and thermal images were also collected using drones from the producer field at Imperial County, CA. The desktop GIS application from Esri (ArcGIS Pro), and python programming language are used to process the satellite data. Similarly, the drone images are processed in Pix4D to create a desired orthomosaic of the study site. c. Summary statistics and discussion of results Preliminary analysis carried out at multiple irrigated fields in southern California using satellite images showed that the tool (VRI-EVAL) provides satisfactory performance in terms of visualization and providing key insights for at-a-glance decision making, and monitor activities. VRI-EVAL is fast, interactive, easily adaptable, user-friendly, and freely available. It displays Google Earth imagery with highlighted areas/fields of interest, along with various elements added, such as, optimal number of zones indicator, gauge showing statistics of selected input satellite product, images of fields arranged into optimal number of zones and performance graph for a selected field. In addition, it gives warning messages, such as "zoning is not required", if the field is almost uniform. Most of the elements of the tool are interactive to each other. The tool will also provide flexibility to the users/farmers to select the feasible number of irrigation MZs by having a look at different available options. The tool works for both circular and rectangular fields representing a wide range of irrigation systems, including center pivot, lateral move, surface, and drip irrigation systems. Outcomes will be validated using the drone images collected and then scaled up to all the fields in Southern California with many more elements and flexibility added to VRI-EVAL. Task 4: Determine the potential adoption of VRI systems in the southern California desert agricultural region through an extensive geospatial analysis a. Major activities completed/experiments conducted To evaluate the potential adoption of variable rate irrigation systems (VRI) in southern California, first, we randomly selected a few center pivot irrigated fields in the region. We then used satellite products with different indices including, normalized difference vegetation index (NDVI), reference evapotranspiration (ETrf), and digital elevation model to delineate the management zones (MZ) of the field. The delineation of MZs was performed using multivariate clustering (k-means) in ArcGIS Pro. After this preliminary evaluation, we selected three different producer fields in Imperial county with the help from the co-PI Dr. Ali Montazar. These fields are planted with forage and vegetable crops. From one of the producer's fields, we collected drone images. We flew DJI Matrice-100 mounted with a multispectral (Micasense RedEdge) and a thermal (FLIR Duo Pro R) camera. Images from this preliminary flight have been cleaned and processed for further analysis. We will use these images, and images collected in the future to validate the results obtained after processing satellite data. b. Data collected Cloud free LANDSAT satellite images for the year 2019 and 2020 were collected from USGS EarthExplorer (https://earthexplorer.usgs.gov/) and METRIC EEFLUX (https://eeflux-level1.appspot.com/) to perform geospatial analysis of the area of interest. We also collected multispectral and thermal images using drones from the producer field at Imperial County, CA. We processed satellite images in ArcGIS Pro using an unsupervised clustering algorithm. The efficacy of site-specific VRI was evaluated using a variance-based evaluation approach. Similarly, the drone images are processed in Pix4D to create a desired orthomosaic of the study site. c. Summary statistics and discussion of results Preliminary analysis done at multiple irrigated fields in southern California using satellite images showed that dividing a field into three to four management zones can address up to 90% of the within-field variability. This shows the significant potential of VRI in the region. Findings will be validated using images collected by drones and then scaled up to all the fields in Southern California. ?A screenshot of the prototype of VRI-EVAL tool is shown below for reference purpose:
Publications
|
|