Aerial photogrammetry to assist farm layout planning and management

Aerial photogrammetry to assist farm layout planning and management

Author: | Date: 18 Jul 2017

CSIRO Agriculture and Food

Take home message

Surface water flows and processes are important factors for production and erosion.  Aerial photogrammetry, by drone or aircraft, can provide detailed information on the soil surface which can be used to inform management.

Aerial photography has been used for many decades to monitor land use and to generate the contour maps of the ground surface. Modern high precision digital photography and computing techniques allow these procedures to be followed with high spatial resolution over larger areas than ever before.


A series of overlapping photographs are taken over the study area so that any single point on the ground surface can be identified in 3 separate images (Figure 1.).  The position of that point is then triangulated in much the same way that the human brain perceives distance to an object based on the views from our two eyes. The demonstration in this paper has applied this technique for every 20cm over an area of over 1300 km2. That is an estimate for over 32 billion locations.  These are then brought together to create a single visual image of the area and a 3-dimensional representation of the ground surface. Testing of this surface indicates that the predicted elevation of each point is accurate to within a few centimetres. This is approaching the accuracy of systems used by surveyors.

A drawing of small aircraft flying overhead capturing 3 images as it advances agricultural field, a roadway and neighbouring woodland.

Figure 1. A point “A” in an agricultural field is identified in three overlapping images. If the position of the aircraft is known for locations 1,2 and 3, the position of A can be calculated. Ground surface points within wooded areas (e.g. Point B) may need to be inferred from other nearby visible points if the view is obscured by foliage.

Once the ground surface is calculated, water flow models can be applied to the calculated soil surface model to show where water is likely to flow.  This is achieved by following water flow paths down gradients from every point in the 3D soil surface.  Furthermore, an estimate of the area of water draining to each point can be calculated as these predicted water flows are followed down through the catchment.

An example showing 4 separate images of the same landscape from a bird's eye view. Four images in sequential order showing steps 1 to 4, step 1 visual imagery from aerial survey, step 2 digital surface model using colour to signify trees and building density, step 3 Ground elevation model using colour to represent elevation and step 4 soil surface water flow model derived from step 3.

Figure 2. The four main steps in deriving the water flow model. 1) Aerial survey conducted using digit photogrammetry. 2) A digital surface model (DSM) is constructed by triangulating the elevation of each pixel in the image.  Note that contour banks and well pads are prominent in the image. The surface also includes surface features such as vegetation and buildings. 3. Surface features are removed and the ground surface beneath them is interpolated. Note that surface depressions (e.g. “Gilgai”) are now revealed from beneath the trees.  4) The ground surface elevation is used to calculate water flows according to small scale topographical variation and features such as gullies, contour banks, drains and roadways.

CSIRO has been investigating the use of these techniques in coal seam gas (CSG) landscapes to provide information on the location and catchment area of water flows for use by land holders and CSG staff during planning for CSG infrastructure placement.  Repeated surveys can show changes in water flow or soil surface elevation which may indicate diversion of water flows, soil loss or build-up of sediment within the survey area.  The likely source of any sediment build-up can be identified by following the predicted water flow paths to that location.  Concerns by land holders regarding surface water flows can be better communicated through the use of a water flow map.  The same maps have been used to communicate the need to locate access tracks along ridge lines or away from significant water flows which may cause ongoing erosion or road damage.

However, such techniques could also be used in many farming systems to study common surface water issues and even simpler approaches using UAVs could prove adequate for such purposes. The following figure demonstrates the ability of photogrammetry to determine the location and size of individual Gilgai within a cultivated field near Condamine, Queensland.  Such information could be used to better understand water flow problems and devise methods for mitigation or remediation.

Screen shot showing Google Earth™ image of a field near Condamine, Queensland gilgai showing superimposed lines representing after rainfall and the corresponding surface water flow model for the same area.

Figure 3. Google Earth™ image of a field near Condamine, Queensland showing gilgai after rainfall and the corresponding surface water flow model for the same area.  Note the model prediction of location, size and shape of gilgai, and more regular flow patterns in areas without gilgai.

Furthermore, such information can be used to diagnose problems with existing water flow control structures and evaluate effectiveness of improvements.  For example, the images in Figure 4 show a paddock with breaches in a series of contour banks (November 2013) compared to the same paddock after work had been undertaken to improve water flows (December 2015).  Changes in surface elevation of the contour banks and resultant changes in water flows are evident in the imagery.

This paper serves to simply raise awareness of such techniques and their potential use by farmers looking to better understand water flows on their farms.  Whilst large scale aerial survey is required to understand water flows at catchment scales, or to understand sources of water flowing across farm boundaries, simple surveys with UAVs may allow digital elevation models of individual paddocks for use in laser levelling or earthworks. However, farmers would be encouraged to discuss this in detail with potential service provides to ensure the level of accuracy required for their purposes.

Four images showing progression between 2013 and 2015. Aerial photos and water flows before and after reworking of contour banks showing improved management of water flows.

Figure 4. Aerial photos and water flows before and after reworking of contour banks showing improved management of water flows.

Acknowledgements

This work has been funded by the Gas Industry Social and Environmental Research Alliance (GISERA) which is a collaboration between CSIRO, Commonwealth and state governments and industry established to undertake publicly-reported independent research. The purpose of GISERA is to provide quality assured scientific research and information to communities living in gas development regions focusing on social and environmental topics.

Contact details

Neil Huth
CSIRO Agriculture and Food
203 Tor Street, Toowoomba Q 4350
Ph: 0417 377 151
Email: neil.huth@csiro.au