Reading the Land from Space — A Compendium on Satellite Imagery and Field Evidence for the Vetiver System
Guide to the seven documents, and the order to read them
This compendium brings together what a land-restoration program can and cannot learn about vetiver from satellite imagery, where the free tools genuinely help, where they mislead, and where the real evidence has to come from the field instead. The worked examples are drawn from Ethiopia – the Witchi, Gebra Dima, Jafti, Mettu, Gule, and Talube catchments, and comparison with the Vanya food forest in India – but the principles, methods, and cautions are general. The resolution limits of free satellites, the discipline of separating what is demonstrated from what is inferred, the difficulty of measuring soil carbon, and the below-ground benefits no camera can see apply to vetiver and restoration work anywhere in the world. Read the Ethiopian sites as illustrations of global lessons, not as a regional case study. A newer strand adds the drone layer that sits between the ground reading and the satellite pixel – potentially the most practical of all, since a low-cost drone serves not only as a measurement bridge but as a working tool for laying out, building and monitoring the vetiver system on the farm itself.
The reading order
The seven documents build on one another. Read them in this sequence: the strategic overview first, then the two how-to companions, then the specific field guide, then the account of the below-ground benefits satellites cannot see, then the note on adding an intermediate drone layer between the ground reading and the satellite record, and finally the cautionary close on soil carbon.
Start here – the strategic overview
The anchor document. It sets out the single idea that governs everything – that the size of what you want to see, relative to a satellite pixel, decides what is possible – and walks through where satellites are genuinely powerful, where they mislead, and where they cannot help at all. Illustrated with real imagery from the program. Read this first; everything else is either a how-to for the techniques it describes or a deeper treatment of a limit it raises.
For anyone new to the tools
The on-ramp. Written for someone with no GIS or remote-sensing background, it introduces the four free tools (Google Earth Pro, Google Earth Engine, QGIS, Python) by what each is for, and lays out a week-by-week learning order where each step produces something useful on its own. Read this second if you will actually be doing the work; skip to the next if you already know the tools.
The consolidated how-to
The reference you graduate into once past the basics. Seven self-contained workflows – digitizing by eye, bare-earth contours, multi-year trend analysis, trend-break testing, sanity-checking a surprising result, hedge-length estimation, and erosion modelling – each with its actual commands and its specific cautions, plus the order they usually run in. This is the practical core for anyone producing results.
A worked field guide
The single most field-useful technique, treated in full: downloading bare-earth elevation data (FABDEM), reprojecting it correctly, and generating accurate contour lines with tree-canopy and building height stripped out – the one output that feeds directly into siting hedges and structures on the ground. Representative of the detailed field guides the Workflows document cross-references; read it when you need to actually make contours.
Where the satellites stop
The counterpart to the whole satellite story. It makes the case for vetiver’s below-ground benefits – the hydraulic function that keeps the root zone alive, and the mycorrhizal, nematode, Striga, stem-borer and termite effects that ride on it – none of which any satellite can see, and all of which matter most exactly where erosion (the visible benefit) is least. Part one is the argument; part two is how to measure each effect in a farmer’s field. Read this to understand what remote sensing leaves undone.
The intermediate scale – and a working field tool
An accompanying scoping and costing note. It addresses the one scale this compendium’s governing idea leaves unreached – the gap between a hand-held reading taken at a single hedge and a thirty-meter satellite pixel. A multispectral drone fills exactly that gap. Flown over a demonstration plot it maps a calibrated greenness signal – NDVI, and over dense coffee or enset canopy the red-edge index NDRE, which holds its resolution where plain NDVI saturates – at centimetre resolution, turning the point-wise ground gradient (root gall, stem borer, coffee berry borer and the AMF proxies read at hedge edge, 2 m, 5 m, 10 m, 15 m and 20 m) into a continuous surface, and over repeat passes on the same RTK-referenced transect into a change-detection series no free satellite can resolve at hedgerow scale. That is the measurement case, and it sits at the Level 2 research tier – best held as a shared instrument rotated across sites, not carried in a farmer’s toolkit.
More important for the program, though, is the far cheaper role the note sets out for an ordinary RGB drone as a working tool on the vetiver farm itself – the point at which the technology has to stay farmer-accessible. A sub-250 g aircraft and the same free WebODM to QGIS chain used elsewhere in this set give a service company, on a first visit, an orthomosaic and a bare-earth surface model of a client’s whole holding: contours to site hedges on the true contour instead of by eye; flow paths, gully heads and ponding areas for a water-first layout consistent with vetiver’s master function; hedge lengths measured off the map and converted directly into slip counts, nursery lead time and a costed quote; and, re-flown each season, a before-and-after record of establishment, erosion reduction and sediment capture that is at once monitoring evidence and the plainest possible demonstration to the next farmer. Read this to see how a drone layer both evidences the gradient and, above all, helps lay out, build and monitor the vetiver system on the ground; the note costs out both tiers – the research-grade multispectral instrument and the low-cost RGB planning drone – in full.
- Vetiver, Soil Carbon, and the Measurement Trap
A cautionary close
A short piece on why soil-carbon sequestration is the most attractive claim a restoration project can make and the hardest to prove, using the Vanya food forest as the worked example of how confident carbon numbers get built on the wrong pool. It closes the compendium on its central discipline: match the confidence of the claim to the strength of the measurement, and reserve strong words like sequestration for figures actually measured. Read it last, as the caution that ties the whole set together.
A thread runs through all seven: satellite imagery is a powerful servant and a poor master. It characterizes landscapes, reconstructs their history, and corrects their terrain – but it cannot attribute cause, it cannot measure soil carbon, and it cannot see the below-ground biology that carries the vetiver case where the visible arguments run out. Used with that discipline, these tools are genuinely valuable to vetiver and restoration work anywhere. Used without it, they produce confident numbers that will not survive scrutiny. The compendium is, above all, an argument for the discipline.