The GEOGLAM-NASA Harvest Agrometeorological (AGMET) Earth Observation Indicators utilize a wealth of Earth observations time-series data to provide valuable insights on in-season crop development and current crop conditions at the sub-national scale. Each AGMET Indicator consists of several EO data plots that quantify critical indicators related to crop health for a specific region and crop over time and are updated every 7-12 days to ensure that users are provided with the most up to date information. Displaying climate, environmental, and vegetative variables that impact agricultural outcomes, the AGMET Indicators were born from an agricultural stakeholder-identified need under the GEOGLAM Crop Monitor initiative for a simple way to visualize and monitor crop health throughout the growing season and with the ultimate goal of identifying potential cropping concerns as the growing season progresses and before a food shortage materializes. Thanks to a collaborative NASA Harvest, GEOGLAM, and University of Maryland effort led by Ritvik Sahajpal (AGMET Indicator visualization development), Antonio Sanchez (website/interactive dashboard development), and Christina Justice (GEOGLAM Crop Monitor for Early Warning Lead), Brian Barker (GEOGLAM Crop Monitor for AMIS Lead), Estefania Puricelli (NASA Harvest Markets & Trade Lead), and Inbal Becker-Reshef (NASA Harvest Program Director and GEOGLAM Crop Monitor Lead), the AGMET EO Indicators are now publicly available for all Crop Monitor regions and crops. The easy-to-use online format allows analysts to select their country, region, year, and crop of interest in order to perform quick and comprehensive crop condition analyses.
While the AGMET Indicators have been used internally by the Crop Monitor team for the past several years, they have recently been expanded to include all the Crop Monitor for Early Warning countries. Given strong community interest in these visualizations, the team is now releasing them publicly through an easy to use online tool. Already, the AGMET graphics are used to support operational monitoring and are frequently implemented as key evidence in several international reports, including the GEOGLAM Crop Monitor monthly bulletins, Crop Monitor Special Reports featuring areas of developing concern, and in country updates from the FAO Global Information and Early Warning System on Food and Agriculture (GIEWS) as well as in support of national analysis in general. FAO GIEWS updates describe anomalous food supply and agricultural situations in order to alert the international community and outline measures to be taken. This is just one use case for the AGMET Indicators, which are included in February’s GIEWS report on erratic weather and conflict conditions in Mozambique, April’s report on drought in the Republic of Angola, and September’s report on food insecurity in Haiti. There is a stakeholder need and inherent value in being able to clearly visualize areas of vulnerability where intervention may be required to mitigate food shortages.
One of the many benefits of agricultural monitoring during the growing season using satellite data is that we can keep track of how well crop production is shaping up relative to previous years in a cost-effective and timely manner. This allows farmers, economists, policymakers, humanitarian agencies, and other agricultural players around the world to more effectively plan, supporting decisions that ensure communities do not go hungry and providing transparency in support of agricultural commodity markets. Put simply, the AGMET Indicators provide satellite derived information to help the agricultural community monitor crop development as it’s happening and enable mitigation measures to be taken early on as problematic conditions develop. It is critical that analysts are able to quickly and effectively compare agricultural metrics over time with previous years as our agrifood system reacts to production amounts that differ widely from what is expected (i.e. the amount of food that was produced in previous years). In order to simplify the complicated interconnected nature of agricultural data and provide a dialed-in view of specific agriculture-heavy regions, each AGMET Indicator is composed of quantitative plots for: NDVI, recent NDVI 5-year comparisons, evaporative stress index, cumulative precipitation (versus the 5 year mean), daily precipitation, surface soil moisture, maximum temperature, and minimum temperature. Ultimately, the AGMET Indicators take the data processing burden off of crop analysts and provide a quick, effective, and easily-digestible means of interpreting massive amounts of global satellite data on croplands as the season progresses.
Each AGMET Indicator plot is crop-specific, meaning that users are able to select the specific crop (e.g. maize, winter wheat, spring wheat, rice, soybean, etc.) that they are interested in analyzing in the specific region that they are evaluating. For the purposes of this section, the elements of each plot are described at a high level. See the “AGMET Indicators: Agricultural Earth Data Applications” section for crop specific examples of how the AGMET Indicators can be used to evaluate a chosen crop in a given region during the growing season.
Within each plot, there are several main components. The light gray area shows the span of the most recent 10-year minimum and maximum values across the season in order to provide a better understanding of the particular EO data product’s variability over time. The black lines show the most recent 5-year mean values across the season in order to demonstrate what average conditions usually represent. The purple lines show what the values were across the previous season. The blue lines show what the values are for the current season in near real-time as the season progresses.
Each plot consists of four vertical lines to provide a reference for where crops are in the development stage, based on the GEOGLAM Crop Monitor crop calendars. The brown dotted line represents the planting stage, the green solid line represents greenup (the beginning of plant growth), the brown solid line represents senescence (late development stage), and the red dotted line represents the end of harvest. By examining the main components of each plot within the context of where crops are in their development cycle, analysts can better interpret how current conditions may impact the remainder of the season and final crop outcomes.
There are two plots for NDVI. The first plot shows the NDVI values for the current season in comparison to the 5-year mean, values from the previous season, and the 10-year minimum and maximum values. As a crop season progresses, NDVI values will increase during early crop growth and development of leafy vegetation. NDVI will then reach the peak as the crop fully develops and reaches maturity, before senescence, and then decreases again as the crop matures and senesces as the leafy vegetation begins to die off. How the NDVI values progress over the course of the season and the individual peak values can help predict the productivity of the crops. The second NDVI plot shows the comparison of the current season NDVI values with the previous five seasons along with the resulting average crop yields from those seasons.
The Evaporative Stress Index (ESI) estimates water loss due to evapotranspiration (loss of water from soil evaporation and from transpiration through plant leaves). The plot shows the current season’s values as compared to the 5-year mean, values from the previous season, and the 10-year minimum and maximum values. Negative values indicate below normal evapotranspiration rates, representative of crops that are stressed due to inadequate soil moisture. ESI is good at identifying soil moisture deficits in the early stages and is able to capture flash droughts, which is the quick onset of drought brought on by periods of hot, dry, and windy conditions.
There are two plots for precipitation, both of which are measured in millimeters. The first plot shows the current season’s cumulative precipitation compared to the 5-year mean. When the cumulative precipitation goes above the 5-year mean, the area between the current values and the 5-year mean is colored in green to indicate above-average cumulative rainfall. When the cumulative precipitation goes below the 5-year mean, the area between the current values and the 5-year mean is colored in red to indicate below-average cumulative rainfall. In addition to recorded precipitation, the chart shows the 15-day forecast precipitation as a dot, either green or red, to indicate whether the forecast will place cumulative precipitation above or below the 5-year mean. The second plot shows the daily precipitation events for the season across time, helping to identify the evenness of rainfall over time or major rainfall events.
During the course of the season, the timing and the amount of rainfall surpluses or deficits can have different effects. As an example, too much rainfall during sowing can prevent farms from planting crops, and flooding events can cause dramatic yield reductions. Alternatively, not enough rainfall early in the season can prevent sown crops from emerging and properly developing.
Surface Soil Moisture measures the amount of available water held in the top five-to-ten centimeters of the soil (depth dependent on soil texture). The plot shows the current season’s values as compared to the 5-year mean, values from the previous season, and the 10-year minimum and maximum values. Up to a maximum of 25 millimeters (mm) of available water is able to be held in the surface soil. Higher soil moisture values during the early development stage help crops to germinate and grow. Mid-range soil moisture values during the season help promote crop growth and root depth. However, lower values typically any time before senescence can cause stress in the crops and stunt development.
There are two plots for temperature, both of which are measured in degrees Celsius. The first plot tracks the maximum daily temperature values while the second plot tracks minimum daily temperatures. Both plots show the current season’s values as compared to the 5-year mean, values from the previous season, and the 10-year minimum and maximum values. Very high temperatures can affect crops in a number of ways such as the increase of water loss through increased evapotranspiration, slow plant growth, reduce yields, pollen abortion, and potentially thermal death. Very low temperatures can also affect crops through the slowing in crop growth, frost or chilling injury, reduced seed quality, and possible yield reductions.
For more information on data usage for crop monitoring including data sources, see the “AGMET EO Indicator Data Sources” section or visit the GEOGLAM Crop Monitor website.
Given the ease of use of the new online dashboard, putting the AGMET Indicators into action is as simple as 1) selecting the country of interest, 2) selecting the crop/season of interest, 3) selecting the region of interest, and 4) selecting the year of interest. This section provides examples of how the AGMET Indicators can help analysts understand crop conditions at a glance, especially in years/regions where food production has been significantly negatively impacted by extreme events (such as drought) or positively impacted by above-average production (bumper crops). When looking at the AGMET Indicators, it is important to keep in mind that the closer we get to the date of harvest, the higher the certainty of the plot outcomes.
During the 2020-2021 winter wheat season that ended in July/August, Ukraine saw a bumper crop harvest (above-average production), despite several other wheat exporters experiencing drought and decreased production the same year. The AGMET Indicators for 2021 Southern Ukraine Winter Wheat clearly demonstrate that conditions were ripe for an excellent production year. For example, the precipitation plot shows precipitation amounts trending slightly above-average compared to previous years, providing much-needed water for crop growth. Looking next at the average NDVI (an indicator of vegetative health), we see above-average metrics throughout the entire season, including in the early vegetative states prior to crop dormancy and more critically during the peak of the season during the reproductive stages. In the Recent 5 years NDVI comparison plot, NDVI is at the highest level in the last 5 years. Overall, the AGMET Indicator points towards a year of excellent production early on in the season and continuing throughout, which were then realized at the time of harvest.
Contrary to the conditions leading to above-average crop production seen in Southern Ukraine, Brazil Maize production in the Mato Grosso Du Sul region had poor conditions corresponding with below-average production. Looking at the amount of precipitation in this region, it is immediately noticeable that rainfall was well below average early on in this region’s maize season and throughout the season. This precipitation shortfall led to the lower than average soil moisture, Evaporative Stress, and NDVI. Together, these AGMET Indicators were able to provide critical and timely information on the growing conditions throughout the season and their impact on crop development, highlighting Mato Grosso as an area of concern to watch closely throughout the growing season.
Drought may have one of the most significant influences on healthy crop production, as demonstrated by its negative impacts on 2021 winter wheat in Afghanistan. The AGMET Indicator for Hirat region in Afghanistan this year started flagging signs of below-average precipitation as early as April, five months before harvest with significant deviations from the expected amount of precipitation from the start of the season. The precipitation plot alone provided strong evidence that Afghanistan winter wheat was experiencing significant drought conditions, and its impact on crop development is clear from the NDVI plot tracking near the 10 year minimum. Knowing that a drought is developing early on in the season and may impact final harvest and ultimately the amount of food that a region is able to produce gives decision-makers time to react, coordinate, and work to mitigate impacts. However, understanding this information and putting it into action requires trusting partnerships, close coordination between organizations, and detailed logistical management.
As data access and technology have made significant advances in recent decades, it comes as no surprise that the amount of available data can often be overwhelming and difficult to decipher. As such, tools such as the GEOGLAM-Harvest AGMET Indicators play a key role for quick and digestible information processing and supporting key agricultural decisions. The AGMET Indicators are currently being processed for each country covered in the GEOGLAM Crop Monitors, which together comprise over 90% of the world’s croplands, with the goal of expanding the tool to include additional countries moving forward. In order to increase food market stability and reduce price volatility, it is critical that market analysts, farmers, and other agricultural stakeholders have a thorough understanding of the amount of food coming to market - whether that be in line with the average amount seen in previous years or more/less. Just as with any other consumer good, supply and demand are the key drivers of agrifood markets and commodity prices. On their own, each plot included in an AGMET Indicator provides valuable information for agricultural analysis, but when taken as a whole, the AGMET Indicators provide much more comprehensive insights on developing crop conditions - not to mention early indication of when and where there might be below-average production or on the positive side, more food production than expected. Satellite data can fill a critical gap in agricultural monitoring, enabling us to not only understand current crop conditions but prepare for potential outcomes in a given growing season, with wider-reaching impacts on market stability, supporting early action and humanitarian response, and bolstering food security.