World food supply amid a pandemic

Part-2: How has the pandemic impacted agriculture activities worldwide?

Sara Aparício
Euro Data Cube
7 min readDec 16, 2021

--

How is the world coping with food supply amid a pandemic?

The World Bank reports that “COVID-19 impacts have led to a severe and widespread increase in global food insecurity”. In a previous article, we gave examples of how complementary satellite Earth observations contribute to monitoring crop conditions worldwide. But crop conditions are only part of the equation. This recent FAO report discusses multiple dynamics of COVID-19 that threaten food security and nutrition: including disrupted supply chains, global economic turmoil, and changes in production.

In this tutorial, we explore the impacts of COVID-19 across some of these dimensions using the Open Source “RACE” and “EO Dashboard” platforms that provide satellite-informed impact indicators.

Satellite observations of COVID-19 dynamics impacting food security

Disruption of food supply chains, loss of income and livelihoods, altered food environments, reduced food productivity and production are examples of how the pandemic affects food systems and food security, and some of the impacts can be observed from space.

Data available on the Earth Observation Dashboard developed by NASA, ESA and JAXA, and the Rapid Action on Covid-19 and EO platform developed by ESA and the European Commission, show analyses derived from satellite Earth Observations that can help us understand some of these dynamics.

Both platforms are based on the Open Source project eodash, available on GitHub: https://github.com/eurodatacube/eodash.

eodashboard.org — The dashboard developed by the three space agencies offers the possibility to track changes in air, climate change, economic activity and agriculture, through the integration of multiple satellite data provided by NASA, JAXA and ESA.
race.esa.int — platform developed by ESA and European Commission providing a wide range of indicators and analytics over European countries, based on data from Copernicus Sentinels, ESA Third Party Missions, Copernicus Services, and other Open Data.

1) Food Productivity and Production — asparagus

In 2019, the Brandenburg region produced about 16% of the German asparagus harvest. The reduced amount of fields cultivated in 2020 was estimated to lead to a significant lower asparagus production and related economic losses for German farmers.

(LEFT): View of eodashboard.org showing how to access the “Productive Area” indicator available for the Brandenburg region. (RIGHT) “Productive Area Change indicator”, exploring the sub-region Brandenburg an der Havel, Kreisfreie Stadt.

Satellite observations — such as from Sentinel-2 allow mapping of cultivated white asparagus fields during the harvesting period showing a significant reduction on area of fields comparatively to 2019. At the same time, the traditional German agricultural labor shortage and the overlap of the first epidemiological wave with the spring agricultural campaign were two additional risk-increasing factors.

The asparagus season started early in 2020 but as soon as the COVID19 lock-down started the cultivated area of asparagus fields was reduced significantly as can be seen in the 20–30% lower area in March and April. Credit: COVID-19 Earth Observing Dashboard, 03–11–2021, Access this indicator: https://eodashboard.org/?indicator=E10a1&poi=DE11-E10a1

2) Altered food environments — rice planting and drought

Satellite imagery from JAXA’s ALOS-2 PALSAR-2 and GCOM-C, ESA’s Sentinel-2 , and NASA/USGS’s Landsat-8 was used to examine the condition of rice plantations near Sacramento, California, USA. These satellites can detect rice during planting, maturing and harvesting by observing land surface conditions such as water inundation for planting and leaf area after emergence.

Rice filed in Sacramento, USA. Credits: Wikimedia Commons

These observations indicate that rice planting, heading and harvesting in 2020 began earlier in many regions than in the past two years, and rice planting area was larger than in 2019. In addition to the impact of weather conditions, there is a possibility that a weather-related increase in rice prices was potentially enhanced by uncertainty over global rice supply due to a potential increase in demand, threats of export restrictions, and labor shortages caused by COVID-19.

NDVI computed from JAXA’s GCOM-C SGLI supports the characterization of planting activity in Colusa, (CA) USA https://www.eodashboard.org/iframe?poi=US07-E10c [ap1] The light/dark red chart background indicates the severity of restrictions or lockdown measures in place at that time, based on data from: Oxford University’s coronavirus government response tracker. Access this indicator: https://www.eodashboard.org/?indicator=E10c&poi=US06-E10c

2021 is expected to be one of the hottest years on record, with precipitation and snowfall well below average. According to the “U.S. Drought Monitor” published by the National Oceanic and Atmospheric Administration and other organizations, large areas in California were affected by a drought of intensity D3 (Extreme Drought) or D4 (Exceptional Drought).

Credit: https://droughtmonitor.unl.edu/CurrentMap/StateDroughtMonitor.aspx?CA

Due to the drought, more than 1,500 dams and reservoirs in California had less than 50% water volume compared to an average year. Due to the severe lack of water for agriculture, the planting area was expected to significantly decrease. According to the “Rice Outlook“ released in July by the United States Department of Agriculture (USDA), the planting area in California was forecasted as 417,000 acres, which is the lowest since 1992. Comparative analyses of ALOS-2 (Synthetic Aperture Radar) satellite data from late May in 2020 and 2021 showed that the area of paddy field filled with water in 2021 was significantly smaller than that of 2020. Therefore, planting area is also estimated to decrease, which is consistent with the USDA Rice Outlook Report.

Comparison result of paddy fields filled with water around Sacramento analyzed by ALOS-2 satellite data on May 28, 2020 and May 27, 2021.

3) Observations vs. Expectations — monitoring row crops harvesting across Europe

The “National Harvesting Evolution” indicator on the Rapid Action on Coronavirus and EO dashboard, is based on the the study “Impact of COVID-19 on Harvest of Row Crops” which quantified when harvest took place in 2020 to understand whether any significant delays were caused due to COVID-19. Observed harvest dates in 2020 are compared to the expected harvest progress in 2020 (simulated within the big data analysis framework of VISTA’s YPSILON® yield prediction service chain, based on an advanced physically-based crop growth model, using natural environmental conditions, e.g. meteorological, Copernicus Sentinel-2 ). The observed harvest progress is derived from Copernicus Sentinel-1 backscatter and coherence time series which are evaluated for winter rapeseed and winter cereals. By inspecting the two curves (radar and model) one can calculate the delay in harvest progress.

(LEFT): Maize. Credits: Schwabe, Kai. (RIGHT): Rapeseed field in Germany Credits: Wikimedia Commons. (RIGHT): Locations where the NATIONAL HARVESTING EVOLUTION indicator is available (at race.esa.int).

Additional information on the restrictions implemented during the COVID-19 pandemic (indicated by the red shade background of the graphs) can help to interpret whether the pandemic had any disruptive effects on the agricultural service chain and harvesting measures.

National Harvesting Evolution indicator https://race.esa.int/?indicator=E10a10&poi=FRb-E10a10

Apart from national level investigations, regional analyses are available on both platform for Germany and Spain.

Locations where the REGIONAL HARVESTING EVOLUTION indicator is available (at race.esa.int).

The regions of Castilla y León and Castilla-la-Mancha are in the top three EU regions for production of winter wheat and barley. Winter cereals are cultivated over around 5.562.000 hectares, corresponding to a production of 16.294.000 tons.

Sentinel-1, Sentinel-2 and Landsat-8 sensors offer a unique dataset to monitor the potential delay or disruption of the winter cereals harvest thanks to the combination of a high spatial resolution (10 meters enabling a parcel-based monitoring) and a very high temporal revisit (up to 2 or 3 days depending on the sensor).

The harvest evolution was monitored, at parcel-level over the whole country. Each week, the percentage of harvested parcels and harvested area has been computed and their cumulated proportion is reported and compared to 2019 for each autonomous region. The 2020 harvesting season started by mid of June which is later than the average crop calendar for winter cereals in Spain.

The assessment of COVID-19 impacts on the 2020 harvest must consider the inter-annual variability caused by the specific climatic conditions (as was the case in 2019). In any case, these satellite based indicators demonstrate existing capacities to monitor the growth and harvest of staple crops at national scale providing timely information on the winter cereal production in an uncertain situation during the COVID-19 outbreak and its upcoming recovery.

This article was co-authored with Anca Anghelea.

More tutorials:

--

--

Sara Aparício
Euro Data Cube

Polar & Space science enthusiast. Earth Observation data scientist at @ESA. Wannabe violinist & northern lights chaser.