For the correct answer to your question, we need to know what remote sensing information do you have? If it's ground RGB photography from drone devices is one thing, but spacecraft - is another matter. For a spacecraft it is necessary to know the spectral and spatial resolution. It is possible to detect the landslide and by circumstantial evidence when the vegetation changes its properties, as beneath the surface of the water, change its direction and can be flooded or dry land.
there is a scientific work of Marco Walter. He investigated the formation of socalled "slidequakes". These are brittle ruptures in sediment caused by starting landslides.
You can use the lidar (terrestrial lidar and lidar UAV). LiDAR not only helps with landslide monitoring but can be used to target potentially susceptible landslide areas. You can also use digital photogrammetric data made from UAV for reducing the cost.
For many landslides in rock, where failure is defined largely by geological structure and fracture network, there will almost certainly be a dilatational phase before final failure. The dilated rock mass will become more permeable and sediments will wash into voids. These can be found and observed by careful investigation. It may take many storms or seismic events to take a rock mass to the critical stage. Surface expression is also of course a clue with cracks developing. In one major progressive landslide (Po Selim Malaysia), there has been about a 16 metre drop in the crest without the full failure developing. This is well described by Professor Andre Malone at HKU who has had students studying it for many years.
Rock movement will also cause acoustic emissions that can be measured. I recall Prof Hoek describing an example from one large open-pit mine where seismic noise allowed the actual failure date to be anticipated. No doubt nowadays careful instrumentation would allow the failure location and size of failure to be identified prior to final detachment.
Re soil-like materials, these often also have pre-failure clues. In a major failure in completely decomposed rock (so weak that it slakes in water), the 20 metre slope was riddled with water channels, many choked with sediments (such as sorted sand) and even graded bedding from sediment deposited into still water (a lake) at 20 metre depth or thereabouts. These had become established over a period of many years before the slope finally failed.
Apart from all this of course there are standard approaches such as inclinometers. and many different types of slip indicators including 'stockings' filled with gravel that rubs and emits noise that can be monitored. People do use face radar to monitor changes but my personal experience on this is rather mixed (even when we knew a slope was moving).
For landslides in soil, the borehole sampling may help with this, when you test the density, moisture content. It would give accurate location of the slip surface, if the inclinometer is installed. For rock landslides, the geophysicsal methods may help.
Steve Hencher's contribution has much to commend it. Getting out into the field is the best way to pick up signs of potential slope failure. Steve outlines them quite comprehensively. I would add that detailed measurements of rock bedding and jointing patterns within the area being studied can pin-point sites of potential failure and indeed the most likely plane of failure. Such work was done along a 2 mile stretch of the A82 road by Loch Ness a few years back and permitted remedial work to be undertaken to stabilise the slopes above the road.
Could I suggest the following publication: 'Slope Tectonics', ed. M. Jaboyedoff, Geological Special Publication 351, 2011.
On slopes where debris deposits or talus are dominant, it is worth following the source areas of pre-existing slides/flows in order to check the depth at which failure occurred. This would identify the depth and nature of any potential slip surface across the slope on either side. It also allows you to examine the sediments in depth, their nature and structure. Following the contour out from such a source area may also reveal tension cracks and similar phenomena as Steve has described.
Dr. Steve has very elaborately explained the phenomenon.
Landslide occurs when stability conditions of the slope is disturbed either by the increase of stress imposed on the slope and / or by the decrease in strength of the earth material building up the slope and it involves enmass downward movement of earth material under the influence of gravity. Determination of causative factors of landslide in any given area will help in demarcating the landslide prone zones. Landslides can be triggered due to external causes or internal causes.
Natural factors which are responsible for landslides to occur are :
1. Presence of faults, joints, bedding planes, cleavage etc., and their orientation.
2. Material properties such as compressive strength, shearing strength, etc., of earth material.
3. Unfavourable ground Slope.
Considering all the above and the past incidences of landslides in and around the area one has to prepare landslide prone areas. These areas have to be continuously monitored using inclinometers or micro-seismometers. However it is difficult to predict accurately when a landslide will take place and also the enormity of the landslide. One of the precautionary measures which can be taken is to move the human habitation from the already identified landslide prone areas during monsoon season.
The best and accurate method is installation of inclinometer through borehole. It will give the accurate failure depth as well as the displacement rate
if boreholes are practical in situ, then the best way to measure subterranean soil deflection prior to a slip is to insert fiber optic in very small borehole flooded with flexible compound to stabilise it (the fiber). the subsequent deflection of the soils at different depths are easily measured with a fiber optic tx/rx system which can be deployed remotely powered by solar with comms link to base monitoring station.