Crop mapping webinars | 2020 - 2021
Summary of Crop Mapping Webinar Series: “Implementing the 2030 Agenda for water efficiency/productivity and water sustainability in NENA countries” (2020-2021)
This short Crop Mapping Webinar Series is provided as part of the learning activities of the project WEPS-NENA. The target participants are those interested to produce, analyze, and interpret crop mapping and water productivity outputs. In addition, some of the webinars could be useful for decision makers in the field of water resources management and agriculture. As the webinars include the case studies of crop mapping from the project activities, this webinar series promote the regional knowledge exchange among the project countries. Though the contents are designed for stakeholders of the above-mentioned project, this webinar series is open for anyone interested in the topics. It is recommended to have a basic understanding of remote sensing, QGIS and access to open source data. The learning goals of this series are to present Methods/tools/analyses that may be used in crop mapping. These materials are supported with examples of issues/problems that crop mapping can inform and case studies of problem-targeted crop mapping from different countries. This series consists of seven webinars presented by speakers from various institutions and will be concluded with a webinar looking into future prospects for using remote sensing for water productivity.
Session 1: Crop mapping using remote sensing
In this webinar, Michela Marinelli from FAO introduces the WEPS-NENA project as well as the four major expected outputs that involve establishment of a robust water accounting system, increasing the crop water productivity, applying the Nexus approach and dissemination of the identified strategies & achieved results. The structure, main features, manuals and trainers of the crop mapping training program are described, followed by crop mapping platform. The methodology of the crop mapping approach is also explained with a guide on downloading, pre-processing, classifying, and validation of satellite images. The webinar concludes with examples of open source tools for image classification such as QGIS and Google Earth Engine (GEE). In addition, examples of open source tools for land use/cover assessment such as WaPOR and Earth Map are also given.
Session 2: Collaborative Crop Mapping of the Urmia Lake Basin, Iran
In this webinar, Mohsen Azadbakht from Shahid Beheshti University introduces the study area, target crops and the field data collection method of the Urmia Lake Basin. The validation of the field data is also described in addition to challenges concerning the field data collection. Some of these challenges involved possible geometric errors in collecting samples and Single & double cropping recognition. The time period was divided into First & Second, according to the order of cultivated crops. The results of the crop map of the first time interval found that irrigated wheat and barley as well as apple, pear and alfalfa comprised the biggest share of cultivation. The results of the crop map for the second time interval showed that non-cultivated areas were dominant in the Urmia Lake Basin.
Session 3: Crop Mapping in Qazvin Irrigation Network
In this webinar, Dr. Bijan Nazari from Imam Khomeini International University and Peyman Jafary from FAOIR highlighted the importance of crop mapping for the Qazvin Irrigation Network and introduced the agricultural profile of the network including its main crops during the different seasons and garden products. The methodology and activities of the crop mapping process was explained which included: a literature review and analysis of available information in the study area, developing the methodology based on the application of pixel-based and object-based classification approaches and Random Forest and SVM classifiers using Sentinel-2 and Sentinel-1 images, preparing data collection instruments and forms, conducting training workshops for field work team, data collection during three different periods of time, analysis and preprocessing of the samples, preparing crop mapping code in Google Earth Engine (GEE), implementation of the methodology, accuracy assessment of the results and finalizing seasonal crop maps. Finally, two consistent seasonal crop maps for Fall and Spring crop seasons in 2019-2020 crop year were presented with high accuracies. The outcome of the results were given and has shown that the developed methodology and GEE codes could be used for crop mapping in QIN and other regions in future if required field data be collected and some minor changes be made based on the characteristics of each study area. Moreover, recommendations were given that using Random Forest classifier, object-based approaches, time series of spectral indices and combination of different datasets lead to improvement in the accuracy of crop maps.
Session 4: Crop Mapping using Remote Sensing in Algeria
In this webinar, Nadia Ait Alioua, Lydia Chaou from the Technical Institute of Field Cultures, Samir Merdaci from the Technical Institute for the Development of Saharan Agronomy and Kamal Kichou from the National Office for Irrigation and Drainage introduces the mapping team by first showing Algeria’s water profile and its water stress situation as well as the role of the SIDA project. The study area of the Hamiz Perimeter was presented which includes its geographical profile as well as the existing water resources. The methodology was provided showing the planning, preparing and collecting of field reference data from the field in addition to displaying the Post-treatment Phase. Overall, it was concluded that the results obtained from Earth explorer remote sensing tool was found to have an overall accuracy of 57,620420081 percent.
Session 5: Presentation of crop mapping by remote sensing in the governorates in Kairouan and Jendouba
In this webinar, Ikhlas Dhaouadi from the Tunisian Ministry of Agriculture, Water Resources and Fisheries presents the two sites of the Kairouan and Jendouba by introducing the irrigated perimeter of Ben béchir, sabaa fyoudh, and Chbika. The methodology was explained showing how the open source data was extracted via sentinel two satellite then subjected to pre-treatment via QGIS preprocessing, and data collection via Open data kit (ODK Collect). This was followed by showing how field data was used to make supervised classification. Afterwards, an evaluation of classification results was presented by analyzing spectral signatures and accuracy parameters. It was concluded and recommended that classes and sample sizes should be specified before going into the field-Work plan, collect more field data, and to separate the references by agro-ecological zones.
Session 6: Contribution of remote sensing to water management in irrigated agriculture - Case of the Berrechid area, Morocco
In this webinar, Dr. Mohamed Chikhaoui from the Agronomic and Veterinary Institute Hassan II introduces the climate related challenges of the Berrechid area including the yearly water resources, demand for irrigation water, and dam capacities. This is followed by an introduction to the crop mapping team as well as the profile of the study area. The methodology is explained by describing the data collection process and data processing steps. The results obtained have concluded that maize and potato comprise the majority of cultivated land in the study area. The results obtained have concluded that maize and potato comprise the majority of cultivated land in the irrigated study area. It was recommended that more data should be collected for more seasons.
Session 7: Crop maps for water productivity assessments using WaPOR
In this webinar, Livia Peiser, Michela Marinelli, Alberto Tordesillas, and Andry Rajaoberison (FAO) introduce the reasons as to why crop mapping is needed and what is available in WaPOR levels 1, 2 and 3 such as actionable near real time information. WaPOR is briefly explained as well as why it is needed and the technical approach it employs given with an application from the Bekaa valley in Lebanon. Earth Map (Google Earth Engine) was also introduced and its use demonstrated. This was finalized with a case study of the Mejerda river basin in Tunisia that starting from the WaPOR land cover map, presented the results of the analysis of the crop seasonal WaPOR data (actual Evapotranspiration, Biomass, Water productivity, Yield and Yield gap) computed using the python scripts available in the New Open Courseware Water Productivity and Water Accounting using WaPOR.