Jason P. Ackerson
Assistant Professor of Agronomy
Office: Lilly 3-317
Area of Expertise: Soil Science
Dr. Ackerson's research and extension program focusses on understanding the spatiotemporal
dynamics of soil physical and chemical properties and developing systems to
utilize this information to improve management of soil systems for improved
agronomic output and ecosystem function.
His research focusses on two main pillars 1) developing and testing
novel soil sensing systems for improved measurement of soil properties and 2)
developing systems to integrate novel sensor information into spatiotemporal
soil mapping and monitoring applications. His extension program focuses on
information and technology transfer; translating data products and tools from
the research sector to the commercial or public arenas.
Low-cost and high-resolution soil data in needed for modern
soil monitoring and mapping applications.
One avenue to collect this data is through the use of proximal (i.e.
near or in contact with the soil) and remote (i.e. UAV, airborne, or
satellite-based) sensors. These sensors
can generate large quantities of data and provide improved data for soil
mapping and monitoring applications across a variety of scales. This sensor data can be used to supplement or
replace costly traditional soil sampling which in turn can improve the accuracy
of soil monitoring and mapping applications.
Dr. Ackerson’s research in this area focuses on the development and
testing of new proximal sensors and developing novel uses for existing remote
sensing products (e.g. satellite imagery).
Soil mapping and monitoring
Soil physical, biological, and chemical properties can vary
greatly in time and space. This
variability can pose a challenge to managers at a variety of scales. For instance, at the field scale a farmer
needs to manage field-level changes in soil properties to maximize agronomic
profitability while minimizing environmental impact. At the regional or watershed-scale,
conservation and land use planners need to understand the spatial distribution
of soil properties to develop effecting conservation interventions and monitor
the success of these plans. Regardless
of scale, understating the spatial distribution of soil properties and
functions (and the associated ecosystem services of these properties). Dr. Ackerson work focuses on utilizing new
and novel data sources to map and monitor the spatiotemporal distributions of
soil properties across a variety of scales.
In-situ measurement of
soil properties with high depth-resolution multisensory penetrometer platform
This project focusses on sensor development and testing of a
multisensory penetrometer platform. The
penetrometer can be inserted into the soil using a hydraulic soil sampling
probe (Fig. 1a). The probe is equipped
with various sensors including a visible near infrared (VisNIR) spectrometer,
electrical conductivity array, and insertion force sensors. Theses
sensors measure soil properties in situ and
in real time at high (2-cm) depth resolution (Fig. 1b). The penetrometer is capable of measuring soil
organic carbon content, particle size (i.e. sand, silt, and clay content),
cation exchange capacity, and soil water content. By collecting soil data in situ, the penetrometer system
minimizes the need for soil sampling and traditional laboratory analysis. This instrument can provide high-throughput
data collection for digital soil mapping and soil assessment. On-going research projects with the
penetrometer platform include:
- Evaluation of the effectiveness of the probe to detect changes in soil physical properties (e.g. porosity, penetration resistance, soil water holding capacity, etc.)
- Development of error-detection routines to determine real-time data acquisition errors.
- Sensor fusion to combine VisNIR and electrical conductivity data for improved sensor performance.
- Improved VisNIR library selection from a priori ancillary data.
Figure 1. The multi-sensor penetrometer platform deployed
in the field (fig. 1a) and an example
high-resolution soil profile measurement
generated using the penetrometer platform.
Informed soil sampling
strategies for precision digital soil mapping
Soil sampling is often the most costly component of a soil
mapping project. To reduce the cost of
sampling and take full advantage of limited resources, efficient soil sampling
routines are needed. While many
strategies exist, few take into account the a
prior information imbedded in existing soil maps. This project focusses on the developing and
validating a soil sampling scheme that uses machine learning to discover and
formalize the embedded knowledge in existing soil maps. This embedded knowledge is then used to
determine the most informative landscape attributes to sample and guides
sampling to the most informative locations. Specific goals of this project are
- Improve sampling efficiency through utilizing the embedded-knowledge in soil maps to direct sampling efforts.
- Develop strategies to identify the economically optimum sampling rate from a priori information.
Figure 2. Three-dimensional soil elevation model with optimized sampling locations.
in dynamic soil properties through improved spatiotemporal land use characterization
Many soil properties are dynamic and therefore are not
stable in time. These properties such as
soil carbon stocks and soil health parameters often change due to changes in
soil management systems (e.g. tillage regime, cover crop use, cropping system)
and land use. Understanding and
monitoring the spatiotemporal trends in dynamic soil properties requires
knowledge of the current and historical land use/management. Improving monitoring of land use and
management strategy requires a multifaceted approach that incorporates existing
and emerging data products with smart soil monitoring and spatiotemporal
statistical methods. On-going research
in this area includes:
- Development of automated "windshield survey" techniques for improved calibration and ground-truthing of remotely-sensed land use and management products.
- Developing tools for remote-sensing of cover crop usage
- Assessing various strategies to characterize and utilize spatiotemporal land use data for predicting soil carbon stocks and soil health parameters