This presentation will set the scene for the GEOGLAM session by providing a high-level overview of GEOGLAM community progress and discuss where we are heading in the next decade.

The first GEOGLAM Crop Monitor was launched almost ten years ago in response to the G20 policy mandate to support global commodity market transparency. The mandate has since expanded to include food insecure regions of the world, with the launch of the Crop Monitor for Early Warning in 2016. More recently work has been extended to the co-development of national crop monitoring systems in less developed nations. Now, as GEOGLAM looks to the future, it has implemented community activities focused on capacity co-development, defining a set of Essential Agricultural Variables (EAV’s), and addressing gaps in open access to in situ data. Together our collective action is addressing policy challenges through the provision of improved EO based information for decision makers at multiple levels, understanding that better information results in better decisions.

As we reflect, the first decade of progress has positioned GEOGLAM to make an even greater contribution to some of the major policy challenges of our time. These include the Sustainable Development Goal’s, climate mitigation, climate adaptation, and disaster risk reduction. Consequently, this is an appropriate time to celebrate and take stock of our progress, while focusing on our response to these emerging challenges.

The G20 GEOGLAM Crop Monitor has been a source of open, timely, and science-driven information on global crop conditions in support of market transparency and early warning of production shortfalls since 2013. Starting with the Crop Monitor for the Agricultural Market Information System (AMIS), which monitors crop conditions across the major producing and exporting countries of the world, the Crop Monitor initiative expanded to cover food-insecure regions of the world in 2016 with the launch of the Crop Monitor for Early Warning. More recently the Crop Monitor has been scaled to support regional and national monitoring systems that have been co-developed in partnership with national ministries and regional agencies that now house these programs. In the last ten years there have been significant advances in the field of agricultural monitoring and the transition of research into operational systems directed at decision support. Looking ahead, we are faced with new impending threats to global agricultural production and food security in the form of increasing frequency and severity of weather extremes and changes in weather patterns due to climate change that are poised to exert the greatest impact on regions already vulnerable to food insecurity, increasing conflict and insecurity, and supply chain disruptions. These are challenges that reinforce the need for strong action from the research community to continue to increase the adoption of operational technologies to support the capacity for governments and the humanitarian community to work proactively to mitigate impacts. While much progress has been made there are still information gaps that the GEOGLAM Crop Monitor community are prioritizing as we look ahead to the next decade including improved cropland and crop type maps, ground truth validation, agricultural impact assessments, long lead weather forecasting and extended outlooks, in season yield forecasting, and continued engagement at the national level. Across these research topics, there is an increased emphasis on connecting information from the field to global scales to strengthen decision support and increase the reliability, accuracy and timeliness of agricultural assessments.

GEOGLAM's Earth Observation Data Coordination Team’s efforts have resulted in considerable advances in acquisition, accessibility, and ease of use of satellite data for agriculture monitoring. As policy frameworks have multiplied, and our G20 mandate has expanded, so too have the observation and product requirements to meet the demand for information. Critical to our mandate is the ability to measure state and change of agriculture, as well as to forecast what is yet to come, in the face of climate change, food insecurity, and land degradation. Following the model of the Essential Climate Variables (ECVs), GEOGLAM has developed a set of Essential Agriculture Variables (EAVs) that are driven by user and policy needs for agriculture information. There are 14 “Top Priority User-Facing Variables” (TPUFVs) that are intended for non-remote sensing audiences (policy and decision support), 21 “Supporting Variables” that underpin analysis to generate the TPUFVs, and 6 “ECVs for Agriculture” that are similar to the Supporting Variables but build synergy with the existing ECVs. Outputs of the GEOGLAM EAV development process include a) specifications (definitions, application relevance, spatial unit of product output (e.g. field or administrative unit), frequency of update, and error assessment requirements), b) a process for identifying/labeling satellite-derived products as compliant with GEOGLAM EAV specifications, c) updated Earth observation requirements, and d) an analysis of gaps and barriers to EAV product generation. Gaps and barriers include data acquisition, data access, methodological development, in situ data for cal/val, training/capacity, and computational power.

This talk will briefly outline the specifications and focus mainly on progress to date while illustrating linkages with existing GEOGLAM-contributing projects, including Sen2Agri, Sen4CAP, Sen4Stats, NASA Harvest GLAM, ESA World Cereals, and more. It will also outline linkages to other GEOGLAM efforts, including Capacity Development, Computation/IT, and In Situ Data Coordination, and provide the vision for the next 10 years of GEOGLAM’s EO data coordination work.

CropWatch Cloud is for multi-scale crop monitoring platform. It provides cloud-based service including modules, area of interest, indices, user interface customization for users and stakeholders by encapsulating and publishing the CropWatch indicators, method and model as web application programming interfaces (APIs) can be composed to execute user specific analyses. Analyses are automatically parallelized to run across many CPUs in the Ali cloud, drastically decrease the amount of time necessary to complete the computations. The CropWatch Cloud allows users to set up their own defined project/system by invoking available APIs including both algorithms and indicators directly to do crop monitoring for any regions of interest using a web browser or local integrated development environment.

CropWatch has successfully carried out capacity building on crop monitoring through module customization in Mozambique, Mongolia and Southeast countries. All indicators in CropWatch were customized, calibrated and made available from a district level of Mozambique. The Mozambique National Agro-Meteorological Bulletin has successfully incorporated the monitoring results from CropWatch Cloud on provinces and districts levels since June 2018. Cambodia and Viet Nam use also CropWatch for their own regional and national crop assessments. CropWatch also automatically feeds information to Thailand Agri-MAP. The drought monitoring functions of CropWatch has been customized as a ten-indicator DroughtWatch system in 2014 for Mongolia, who regularly disseminate drought monitoring bulletins to government bodies. DroughtWatch has also been localized and used in Cambodia, Myanmar and Sri Lanka. Recently, capacity building actions are carried out for 14 countries over Africa and Asia.

Supported by the data and information provided by CropWatch APIs, CropWatch Cloud also offers a website for users all over the world to make their own analyses, which greatly enhances the transparency of crop monitoring information. CropWatch also provides a timely update on crop information and fills the information gaps in developing world to support climate adaptation and policy making globally. CropWatch Cloud is an integrated platform designed to empower not only the common remote sensing scientists, but also a much wider user communities globally that lacks the technical capacity to conduct crop monitoring, especially for the developing countries.

The Asia-RiCE initiative (http://www.asia-rice.org) has been organized to enhance rice production estimates through the use of EO, and seeks to ensure that Asian rice crops are appropriately represented within GEOGLAM. Asia-RiCE is composed of national teams that are actively contributing to the Crop Monitor for AMIS and developing technical demonstrations of rice crop monitoring activities using both Synthetic Aperture Radar (SAR) data (Radarsat-2 from 2013; Sentinel-1 and ALOS-2 from 2015; TerraSAR-X, Cosmo-SkyMed, RISAT, and others) and optical imagery (such as from MODIS, SPOT-5, Landsat, and Sentinel-2) for 100x100km Technical Demonstration Sites (TDS) as a phase 1 (2013-2015) and wall-to-wall (2016-2018) in Indonesia, Vietnam and Cambodia in Asia with satellite –based cultivated area and growing stage map. The Asia-RiCE teams are also developing satellite-based agro-met information for rice crop outlook, crop calendars and damage assessment in cooperation with ASEAN food security information system (AFSIS) for selected countries (currently Indonesia, Thailand, Vietnam and Japan; http://www.afsisnc.org/blog), using JAXA's Satellite-based MonItoring Network system as a contribution to the FAO AMIS outlook (JASMIN) with University of Tokyo (http://suzaku.eorc.jaxa.jp/cgi-bin/gcomw/jasm/jasm_top.cgi). This paper describe the current status of Asia Rice team activity.

We will present Digital Earth Africa’s effort to produce a cropland extent map for the African continent and other Earth observation based tools and products to address food insecurity across Africa.

Digital Earth Africa (DE Africa) operates a digital infrastructure in Africa that aims to provide free and open access to Earth observation (EO) data and services to all, and build capacity across Africa to use EO based insights to address sustainable development challenges. The range of data and services provided by DE Africa, including analysis-ready surface reflectance, surface temperature, and synthetic aperture radar time series will support many types of crop monitoring applications. Our free analysis environment allows any user to access data through an Open Data Cube API and prototype applications before scaling up to regional or continental services.

The 10m resolution cropland extent map is DE Africa’s first continental service for agriculture and is expected to serve as a basis for higher level crop monitoring and management products. We have co-developed this service with our partners across the continent, including initial consultation and validation supported by GEOGLAM. Members of the regional African geospatial institutions (RCMRD, OSS, Afrigist, AGRHYMET, and NADMO) were instrumental in defining the specifications of the product, in developing and implementing a continental scale reference data collection strategy, and assisting with iterative model building.

We have classified cropland using an annual Sentinel-2 time series and a Random Forest machine learning model. The product comes packaged with three layers: a pixel-based classification, a pixel-based cropland probability layer, and an object-based segmentation filtered classification. All the components of the service: models, reference data, code, and results are open source and freely available online through Digital Earth Africa's mapping and analysis platforms.

Driven by the needs of the African community, DE Africa strives to provide better tools and services to support agriculture and food production in Africa. By early 2022, we expect to provide an operational monthly NDVI Anomaly service based on the Landsat and Sentinel-2 archives. We will start to develop methods and tools for crop type classification and will present our early results. We will also present our efforts on the collection and sharing of in situ and reference data to support application development.