A Quick Comparison of FEWscapes Models with Common Land and Water Management Models

When looking to the future, when the world will be very different than today, one thing that can be certain is the basic principles and processes of nature – e.g., photosynthesis, nutrient cycling, gravity – will work the same. That is why process-based models, which simulate these basic natural principles through a series of equations, work well for understanding possible outcomes of climate, land use, and policy changes on future food, water, and energy systems.

The two primary models used in FEWscapes are process-based. Agro-IBIS simulates the biogeochemical processes associated with the production and management of major cropping systems, while THMB is a routing model that simulates how nutrients and water flow across the landscape.  

In FEWscapes, we are combining Agro-IBIS and THMB to simulate possible changes in climate, land use, and economic policy over the next three decades in the Upper Mississippi River Basin. As a pair, Agro-IBIS and THMB can estimate what such changes could hold for the future state of ecosystem services – such as crop production, water quality and quantity, and flood regulation – that allude to the future state of food, energy, water, and ecosystem security.

But how do Agro-IBIS and THMB compare to other models commonly used in land and water management in the Upper Mississippi River Basin, such as EPIC, SWAT, SPARROW, MODFLOW, and SnapPlus?

In November of 2021, we hosted a Lunch and Learn about how they compare for our community partners. While I won’t compare them all in this post for the sake a brevity, I will provide a quick comparison with SPARROW and SWAT, two that are particularly well known among researchers and managers who work on nutrient loss reduction, an issue that influences food, energy, water, and ecosystem security.

If you want the full comparison, I recommend watching the recording of that Lunch and Learn at the end of this post. The following chart provides a snapshot of the comparisons. You’ll likely need to know a little bit about modeling to decipher it.

First, to be clear, in comparing these models, we aren’t suggesting a battle of the models. When it comes to choosing a model to inform land and water management decisions, it is a matter of goodness of fit, based on the goals and priorities of the exercise. Each model has its strengths and limitations.

SWAT has perhaps the most similarities with the Agro-IBIS and THMB duo. It too is a process-based, hydrologic and biogeochemical routing model that researchers and managers use to answer questions about water quality and quantity and the environmental impacts of land use and management. In fact, it is sort of also a paired model, as another process-based model that is similar to Agro-IBIS, called EPIC, is integrated into it.

Also similar are the types of data that go into SWAT – e.g., weather and climate, land cover, basin boundaries, and nutrients from fertilizer and manure – as well as the outputs, such as crop production, nutrient loads, and streamflow. Calibration works pretty much the same way too, using field measurements and monitoring data to validate the model.

Perhaps the biggest difference is that SWAT can account for more detail when it comes to irrigation sources and conservation agriculture practices, such as cover crops and low-or no-till, on the landscape, while Agro-IBIS and THMB don’t get as detailed about those factors. Minor differences include SWAT’s ability to account for pesticides and bacteria in water, which Agro-IBIS and THMB don’t do, and the resolution of their outputs – where SWAT uses the HUC watershed scale, Agro-IBIS can produce results at a smaller scale. For instance, in FEWscapes, we’ll be simulating a scale of one square-kilometer.

Agro-IBIS also simulates the transport of water and energy through the soil and plant canopy in a much more detailed way compared to SWAT. This means it is able to represent processes like photosynthesis and soil hydrology in perhaps a more realistic way and allows for more complex analyses of how factors such as carbon dioxide and moisture impact plant growth over time.

Another model that is critical to the nutrient loss conversation in the Mississippi River Basin is SPARROW, but it is quite a bit different from Agro-IBIS and THMB in terms of how it works and what results it can produce. SPARROW is focused squarely on estimating nutrient loads to determine water quality, rather than measuring a variety of ecosystem services like Agro-IBIS and THMB.

Also, where Agro-IBIS and THMB can calculate changes in nutrient loads across a variety of timescales – from daily to yearly – SPARROW looks only at long-term mean annual loads of phosphorus and nitrogen. It also uses very different means to calculate these outputs – a mix of statistics and mass balance equations, rather than process-based equations – which requires a different way to calibrate the model.

Furthermore, SPARROW calculates nutrient loads based on current conditions, which doesn’t make it suitable for use with future scenarios; although, there have been efforts to expand its scope to be more forward-looking.

All this said, SPARROW achieves its main goal: to use the extensive water quality monitoring network of the U.S. Geological Survey to provide a statistical basis for estimating nutrient loads in places that don’t have monitoring, and to give a yearly snapshot of nutrient loads in the Basin and beyond. We encourage you to check out the USGS SPARROW Mapper for some cool data visualizations of nutrient loads in the country’s major regions.

There are always tradeoffs in model selection – no one model will do everything. For example, models that use statistics and empirical data to estimate outputs, such as SPARROW, work really well in decision-making realms, because they don’t require as many data inputs and can run fairly quickly. In comparison, process-based models like Agro-IBIS and THMB often require many datasets and can take a lot of time to run to get results…a time lag that can be a drag on decision-making.

However, one tradeoff of that time lag with Agro-IBIS and THMB is their capacity to step more incrementally through time, day by day, into the future. This provides an opportunity to assess the effects of short-term events like floods and droughts that could have long-term impacts on food, energy, water, and natural systems.

If you’re intrigued to get a more detailed comparison of these models, watch the following recording of the Lunch and Learn led by FEWscapes modelers Eric Booth and Kelsie Ferin.