Crop Insurance Today Fourth Quarter 2022 - 26

as reflected on the visible and near-infrared wavelengths
of spectral data. Radar-based satellites
measure the moisture content in the topsoil layer
as reflected on longer wavelengths (microwave) of
the spectral data and have the additional advantage
of working when it is dark or there is cloud interference.
Both types are available at varying degrees
of resolution and accuracy. Optical satellite technology
is used in vegetation index type products
such as NDVI (Normalized Difference Vegetation
Index), while radar-based satellite technology is
used in soil moisture index type products.
Various applications were presented, including
NDVI-based tools and products for customers,
satellite data for loss adjustment, machine learning
for claim settlement as well as soil moisture index
products. Machine learning involved predictive
modeling based on satellite imagery and was
validated by some ground truth sampling. However,
difficulties were mentioned as part of sampling
for ground truthing purposes. It appeared
that the damage assessment based on machine
learning models needed to become more accurate.
In addition, algorithms required extensive
data for training and needed to be recalibrated for
each year and location. One of the soil moisture
index products was targeted for drought peril and
got started this year in Romania.
Another application involved a hybrid model
combining remote sensing and predictive yield
modeling.⁶ This approach relied on the zoning
of rice ecosystems in setting insurance threshold
yields for rice in Cambodia. It was intended for
providing insurance solutions for small-holder
farms, which has been an ongoing challenge.
Satellite data-based insurance is viewed as
an alternative to weather index products. One
application of the latter is on developing grassland
yield insurance in Germany. Using weather
(precipitation, temperature, etc.) and management-related
data, the product is designed to
measure yield as well as the quality of forage. This
⁶ This is part of RIICE (Remote sensing based Information and Insurance for Crops in emerging Economies) project, which is undertaken by a
partnership between insurers, governmental organizations, an research organizations. For more information visit http://www.riice.org/.
⁷ For a detailed discussion on the said sources of basis risk, see Geoville (2022, p.5) as referenced in footnote 5.
is a nascent product attempting to overcome the
challenges of measuring yield in forage.
Some key challenges remain with index insurErrors
and Omission Insurance
For Your Agency
Full lines of coverage including MPCI Crop Insurance
We will work diligently to offer you quotes with
reputable companies at competitive prices
To obtain a quote for your agency call 1-800-769-6015
American Insurance Services, LLC.
Premium financing is available
We have over 35 years experience in
all lines of insurance
www.t omsta n l e y ins ur a nc e .c om
24 FOURTHQUARTER2022
ance in general and satellite-data-based index insurance,
in particular. Chief among these challenges
is the basis risk, which is essentially the degree
of mismatch between farm-level losses and the
indemnities generated from index insurance. The
basis risk may stem from spatial variance, sensor
instrument, remote sensing technology, procedural
(data collection related) factors, product design,
temporal challenges, and missing data substitution
methods.⁷ Data integrity also poses an important
challenge. Satellite imagery-based data need to be
validated through some ground truthing and the
underlying models need to be calibrated. Farmers
may not trust satellite-data-based products otherwise,
or the effects of such products can be too
convoluted for them to discern. Crop-type identification
is another challenge. For instance, similar-looking
cereal crops such as wheat and barley
cannot be differentiated at the typical resolutions
of satellite imagery. Finally, very high-frequency
and resolution satellite data come at an increasing
cost to insurers. Interpretation of vast amounts
of satellite-based data practically and effectively
requires investments in additional analytics and
information technology infrastructure.
While satellite technology-based products carry
considerable promise, farmers' eventual uptake
of these products remains to be seen. It may be
prudent to remember that market penetration of
index products has been low around the world.
For instance, the market share was less than two
percent in Latin America despite it being a decade
into their introduction there. The grassland
insurance products were already offered in several
countries in Europe, but the uptake has been on
the low end so far. The PRF-VI (Vegetation Index)
product was discontinued in the U.S. in 2014
because of a lack of interest. On the other hand,
the premium volume for PRF-RI (Rainfall Index)
product in the U.S. had an increasing trend after
the 2014 Farm Bill and exceeded $1 billion in 2021.
Such a differential uptake between the two PRF
products may be explained by farmers' perception
and trust in the product as mentioned earlier.
Conclusion
The diverse and interesting program of the
36th Congress of AIAG facilitated a valuable
exchange of information among international
experts. As agricultural risk insurers and farmers

http://www.riice.org/ http://www.tomstanleyinsurance.com

Crop Insurance Today Fourth Quarter 2022

Table of Contents for the Digital Edition of Crop Insurance Today Fourth Quarter 2022