VanderSat

Toensureearlyandaccuratein-seasonyieldforecasts,anywhereontheglobe

By comparing VanderSat soil moisture, vegetation optical depth, and land surface temperature, with historical yield data, we can determine the impact of water supply & demand on local crop yields throughout the growing season, and use these relations to forecast crop production via a machine learning procedure.

Our long historical satellite data record provides a very large amount of training data, which ensures that our cross-validated prediction models perform well one growing season after another. Provided that local yield data is available, this approach ensures that early and accurate in-season yield forecasts are available anywhere on the globe.

Develop your own forecasting service

Replace expensive field observations with an automated forecasting service, which uses our unique remotely sensing data sets and machine learning approach.

Tailor-made input parameter sets for your models

Improve your own forecasts with VanderSat’s unique, industry-leading soil moisture, temperature, and vegetation health index parameter sets, delivered daily, and tailored to your existing model.

Soil moisture monitoring service

Get early insight into the water availability of the crop in order to make more informed trading decisions, daily and near real time.

Competitive edge

Accurate and timely crop yield forecasts are essential to global food security, so a substantial amount of research has been carried out over the years in generating such prediction models. However, it remains challenging to understand the influence that precipitation has on crop yields since it is merely an indirect measure of a plant’s water availability. Therefore, most models instead heavily rely on one public datasource parameter, namely NDVI. The NDVI parameter, however, is hindered by clouds and, when it comes to crop health, only captures the symptoms, and not the root causes.

Here at VanderSat we have three unique satellite products at our disposal, all derived from cloud-penetrating passive microwave satellite data (soil moisture, vegetation optical depth, and land surface temperature), with long historical records (2002 -), and, thanks to our patented downscaling algorithm, available at an unparalleled 100 m x 100 m spatial resolution. Each of these directly measured metrics interacts differently with each of the crop development stages during the crop cycle: together they can provide insight into crop progress throughout the entire growing season at a high spatial resolution.

Our forecasting methodology also allows for fully scalable yield forecasts: we can deliver crop yield forecasts either on county or municipal scale, a regional scale, or on a national scale. The global availability of VanderSat datasets ensure that yield forecasting services can be set up to support your projects anywhere on Earth.

Current Data Landscape

NDVI

Hindered by clouds and darkness.
- Less data points, so when you actually need it, it could be that there is no measurements for a month.
Not an indicator about available water for the plant, but a results of that. In other words, it show symptoms instead of root causes.
NDVI data sets are publicly accessible and are used by all market players; limited competitive advantage.

Rainfalldata

The quality of these data sets vary strongly in different regions.
Often low quality and inaccurate data, that will influence crop models in a negative way.
Indirect measurement. You need to understand the effects of rainfall intensity, direct runoff (water that goes into the river streams running over land) and interception (water that is caught and evaporated from the vegetation)?
Data sets are publicly accessible, limited competitive advantage.

Soilmoisturesensors

High costs per sensor.
Labour-intensive e.g. placing each sensor and calibrating.
Historical data can be available for several years. However the quality of multiyear observations are doubtful.
- No climatology and indication if situation today is drier or wetter than normal.
Point measurement is not able to monitor large producing regions and countries.
Not scalable.

VanderSat is a leading provider of global satellite-observed data, products and services over land with a special emphasis on water and crops. We’ll work together to define the scope and parameters of the forecasting service to best suit your specific crop, region of interest and trading strategy.

Currentservices

The annual variation of U.S. corn and soybean yield has implications for global food security. The country ranks first as the largest corn exporter in the world and second for soybeans. As demand for food, feed, and fuel continues to rise, unexpected declines in the production of the U.S. can lead to food shortages and rising prices. Therefore, forecasting crop yields can provide an early insight into a crop’s supply, and with that fluctuation in price.

Our current U.S. forecasting service, based on only our unique satellite data, is able to predict these annual yield variations earlier and with far more accuracy than the United States Department of Agriculture. As a result, our yield forecasts have proven to be a profitable trading strategy earning an annual return of 13% from the years 2002 till now.