Conference Agenda

07-03: Applications of earth observation in rural areas
Wednesday, 27/Mar/2019:
2:00pm - 3:30pm

Session Chair: Thomas Esch, DLR, Germany
Location: MC 2-800

ID: 599 / 07-03: 1
Individual Papers
Topics: Use of remote sensing and land use policy
Keywords: Large Scale Land Acquisitions, remote sensing, monitoring

Large-scale land acquisition monitoring with high resolution imagery retrieval and profiling in the ASAP platform

Felix Rembold1, Guido Lemoine1, Matthias Hack2, Christof Althoff3, Patrick Griffiths4, Ferdinando Urbano1, Gabor Csak1

1Joint Research Centre of the European Commission, Italy; 2Deutsche Gesellschaft für Internationale Zusammenarbeit, Germany; 3GIGA German Institute of Global and Area Studies, Germany; 4European Space Agency, United States of America

Detailed geographic information on Large scale land acquisitions (LSLA) in developing countries is generally not easily available due to several reasons, including low transparency of such deals, remoteness of the areas concerned and conflicts about tenure rights. In such a situation remote sensing is one of the most promising means for mapping and monitoring LSLAs during their implementation, by detecting land cover and land management changes visible from space. The high resolution viewer of the ASAP (Anomaly Hotspots of Agricultural Production) platform is a recently developed example of such an application. High resolution image time series visualization and analysis provides the geographic evidence allowing detection of phenological crop stages, large open fires, vegetation clearing, flooded areas or new infrastructure (eg. irrigation, greenhouse, roads) implemented in each parcel of an LSLA. This ideally complements non geographic information collected by other projects such a for example the Land Matrix and facilitates impact monitoring.


ID: 660 / 07-03: 2
Individual Papers
Topics: Data integration & interoperability for public service provision
Keywords: Geo-spatial, Big Data, Water Resource Management, Indus Basin

Geospatial big data platform for water for all in Indus basin

Ather Ashraf1, Muhammad Abid Bodla2, Ijaz-ul-Hassan Kashif3, Ch. Muhammad Ali Nazir4

1University of Punjab, Pakistan; 2Member Water, Planning and Development Department, Govt of Punjab, India; 3Executive Engineer, Irrigation department, Govt of Punjab, India; 4Assistant Chief (Coordination), Planning and Development Department, Govt of Punjab, India

Indus Basin is the backbone of Pakistan water resources and it plays important role in providing water for drinking and agriculture purpose. Although this basin along with its major rivers and glaciers provide rich resources pertaining to water availability, the lack of water governance turns this blessing to a disaster like flooding and drought. An important step towards attaining water governance is to obtain better water information and sharing. The government has taken a step in the direction of use of "Big Data" with data clearinghouse for hydro-meteorological applications at Indus basin level. Big data is helpful in storing and extracting useful information about water resources which is achieved by analyzing data statistically in the temporal and spatial domain. The platform also uses hydraulic and hydrological modeling and socioeconomic data to analyze vulnerability and risk especially for the marginalized area within the river basin.


ID: 554 / 07-03: 3
Individual Papers
Topics: New ways of land data capture & analysis (incl. machine learning)
Keywords: soil, africa, sensor, test, laboratory

Realtime digital soil fertility data for fact-based fertilizer selection by smallholder farmers

Christy van Beek1,2, Sally Musungu2, Rob Beens1, Angelique van Helvoort1

1AgroCares, Netherlands, The; 2SoilCares Foundation, The Netherlands

Recent technological innovations in IT, sensor technology and machine learning have opened the possibility to use Near InfraRed (NIR) sensors for on-the-spot, real-time and affordable soil tests within 10 minutes using a Bluetooth connection between the NIR sensor and a software application for data interpretation on a smartphone. Within 10 minutes the farmer receives a soil status report and a fertilizer recommendation for his specific crop selection. This innovation was first released in Kenya in 2017 and has rapidly expanded to 15 countries, and growing. In this paper, the innovation and experiences and new developments since the introduction of the innovation are presented. Within 1 year about 25000 farmers were reached of which more than 50% imported significant yield increases, about 75% changed their farming practices and more than 80% requested soil tests for the next season. By integrating soil data into data platforms more holistic interventions can be developed.

07-03-van Beek-554_paper.pdf
07-03-van Beek-554_ppt.pptx

ID: 722 / 07-03: 4
Individual Papers
Topics: Use of remote sensing and land use policy
Keywords: Remote sensing, farming systems, food production, Africa, land planning

New ways to use remote sensing based phenology and machine learning for mapping irrigated and rainfed agriculture in Africa

Tobias Landmann1, Natalie Cornish1, David Eidmann2, Jonas Franke1, Stefan Siebert3

1Remote Sensing Solutions GmbH, Germany; 2Technical University of Darmstadt, Darmstadt, Germany; 3University of Goettingen, Goettingen, Germany

In spite of the need for consistent, explicit and large-scale cropland/farmland information at high spatial resolution for land management decision making and food production estimates, these data sets are not yet readily available for Africa. The cropland layers that are currently available in Africa do not provide thematic detail beyond cropland and non-cropland at a fine spatial scale and essentially do not exploit the wealth of information extractable from longer time-series data (now available from well-processed 30-meter Landsat or 10-20-meter Sentinel time-series data). In our approach, we showed how better thematic detail and mapping accuracies can be attained in mapping irrigated versus rainfed agriculture in Africa using only the function parameters from best fit harmonics, derived from long-term 30-meter Landsat vegetation index observations. We are confident that the current method can be employed for effective and accurate land use mapping and as such complement future land use policy plans.


ID: 774 / 07-03: 5
Individual Papers
Topics: New ways of land data capture & analysis (incl. machine learning)
Keywords: Remote sensing, real time monitoring, geo tagged, area audit, crop growth status

Use of Remote sensing technology in small holder supply chains in Asia

Suparna Jain, Harsh Vivek, Rajpal Singh, Ernest Bethe, Bas Rozemuller, Rahmad Syakib, Krishna Kumar

International Finance Corporation, India

Majority of farmers in Asian countries like India and Vietnam are smallholders, farming on less than two hectares of land. As food demand increases by 20% and arable land keeps getting scarce, yield improvements through smart land-water use management has the potential to increase food availability. Yield gaps exceed 50% in many Asian countries owing to technology gap. New business models in agriculture, leveraging technology through data analytics and artificial intelligence, can help farmers access information related to their land, agri-inputs, weather, finance, and markets, thereby helping them increase yields, improve incomes,resilience and traceability. In this scenario, IFC MAS advisory is working with CropIN and Farm Force in sugar and coffee value-chains of DSCL (India) and Simexco (Vietnam) to deliver:

• GIS and remote sensing solutions for digital monitoring and digital management of 5000 sugarcane and 5000 coffee farms

• Smart weather-risk digital solutions to farmers providing real-time weather forecast and crop-advisory