Blog by Dr. Marcela DOUBKOVÁ
Expert in Remote Sensing, Pessl Instruments GmbH

Farmer’s equipment has grown larger, more intelligent, effective and often also more expensive. To effectively manage inputs, Workforce and state-of-the-art Machinery in changing climate, farmer expects improved understanding of the situation in the field. They require exact spatial information on the accessibility, that would help them decide when and where to start the operations, optimize their sequence, as well as the number of machines in the field.

The Workflow tool within EUXDAT will provide farmers exactly such information on field accessibility. It should be used instead of field visits prior to any operation. Also, it should avoid useless traffic with heavy machines.

The traditional services often tend to see the problem of field accessibility to be dependent mainly on rain (the rain of the day before yesterday, the rain of yesterday and the rain of today). While this may work fine in flat areas, already over small undulating hills the lateral water movement becomes significant and thus the information on the elevation is required to correctly judge about the water distribution.

Workflow will offer a product describing field accessibility for next 5 days by spatially modeling the water movement in your field. It relies on the present in-situ soil moisture data from iMETOS stations (Figure 1) and forecasted weather data from METEOBLUE as well as elevation information from the Shuttle Radar Topographic Mission (SRTM). It has been demonstrated that by deriving simple slope-related indices from the DEM data, one can localize dry shoulders, wet depressions or flow channels that are closely related to soil water distribution (Figure 2). In particular, we use the topographic wetness index (TWI) that represents the possibility of accumulation of water. It ranges usually between 1 – 14; 1 being on steep slopes where water runs off immediately, 14 being on very even regions where water accumulates from neighboring areas (see Figure 2).

Figure 1. Soil moisture levels and precipitation available from the iMetos station from Pessl Instruments in 2019

Importantly, the service gets an advantage of short and near infrared bands available on board Sentinel-2 satellite. In particular, we use both SWIR  bands and NIR bands onboard Sentinel-2 to judge about the soil water and open water bodies. By doing so we thus continuously teach the model to new events. Figure 3 nicely represents the fact that areas below undulating hills accumulate more water shortly after a precipitation event while it becomes relatively homogenous in dry portion of the season.

Figure 3. NMDI water index over the field throughout the non-vegetation part of the year. NMDI ranges between 0.5 – very wet to 1.1 – very dry. NMDI was assessed over scenes with no vegetation.

Our intention is to develop an event-based field specific model capable of simulating a single rain event from existing datasets. The model should be capable of describing the upper few cm movement of water and have an ability to expect ponding in depressions. The field accessibility product will help the farmer or farm manager to optimize the sequence for the week, day or the next task during the work day.