Forest Inventory: Building an elevation surface

Introduction

Elevation data is very useful in forestry and natural resources management, especially in more rugged landscapes. It's a basis for delineating watersheds; generating viewsheds; determining depth to the water table; modeling animal habitats; laying out, designing, and costing roads; and scheduling site-aware harvests and silviculture, to name a few.

Until recently, however, reliable elevation data was difficult to acquire, expensive, and typically not very accurate. Today, accurate digital elevation data is widely available, often for free download over the Internet. Federal, provincial, and state government websites are primary sources of the latter.

Typically, elevation data is made available as spot observations across a geographic area. Most often, these elevation observations are acquired through some form of remote sensing. In the past, laborious aerial photo interpretation methods were often used to capture spot elevations. Today, airborne sensor systems, such as lidar, are commonly used. Regardless, the raw data produced takes the form of x, y, and z observations, termed mass points.

Mass points by themselves, however, are of little use in the aforementioned applications, where a continuous elevation surface, or digital elevation model (DEM), is required. The latter may be implemented as a raster or triangulated irregular network (TIN). You can find much more detail about DEMs at the Wikipedia website, http://en.wikipedia.org/wiki/Digital_elevation_model.

Okay, how does one build a DEM from mass points?

Location

A small 1400 ha woodlot in the Acadian-New England forest region of North America.

Time to complete the lab

Approximately 3 hours

Prerequisites

A basic working knowledge of geographic information system (GIS) technology, and ArcGIS software in particular (ArcScene experience is useful, and previous exposure to the 3D Analyst extension is good too. Familiarity with the Woodlot geodatabase inventory is essential.)

Data used in this lab

A mass points file and a personal geodatabase of several feature classes for a small (1,400 ha) woodlot in the Acadian-New England forest region of North America (All data is NAD83 datum with New Brunswick Double Stereographic projection, unless otherwise stated.)