|
|
| Libyan Arab Jamahiriya Land Cover Mapping |
 |
The land cover mapping of Libya was conducted as one of the
activities of the "Mapping
of Natural Resources for Agricultural Use and Planning in Libya" (LIB/00/004)
project.
Project background
The project was initiated by the Government of Libya, FAO and the
United Nations Development Programme (UNDP) to strengthen the
capacity of the General People's Committee for Agriculture,
Animal and Marine Wealth (GAAAMW) to manage land resources at
national and sub-national levels through the establishment
of a strategy, and a spatially based operational decision
support system - the Land Resources Information Management System
(LRIMS).
Land Cover Mapping
The land cover of Libya was developed as component of the broader
LRIMS database.
The photo-interpretation work has been performed on 88 Landsat
ETM images (180x180 km), 18 covering the Northern area (Phase I)
and the remaining 70 covering the Southern area (Phase II).
The acquisition date of the images ranges between 2001 and 2002.
The images used on the 1st phase of the Project have 15 meters
resolution since the panchromatic band was previously added to the
432 RGB composites.
The mapping scales were 1:50,000 for agricultural areas and other
areas characterized by a mean annual rainfall equal or above 200 mm,
and 1:100,000 for areas characterized by a mean annual rainfall lower
than 200 mm, mainly in the southern part of the country.
The legend adopted for the land cover product of Libya reflects the
requirements of the different regions of the country. For example,
the characteristics of Southern Libya led to the introduction of new
classes to allow the detailed description of bare areas, the
oasis environment, and its surrounding natural vegetation.
The final legend consists of 108 LCCS-based classes. The use
of LCCS in the development of this product ensures the compatibility
with other similar datasets and a broader use by the community. Following
is the summary of the final land cover data set:
- 33,551 polygons covering an interpreted area of 166,560,000 hectares;
- 108 single classes used for the interpretation;
- 755 mixed units deriving from the combination of the single classes;
- 149.6 MB sized file for the full resolution product.
The Libyan full resolution dataset is the most important output of
the project activities. Using this data set, a number of aggregated
products have also been developed to meet the needs of the different
users. For example, one aggregated land cover data set of 10 classes (see below)
was realized:
| Class | Description |
| AGRICULTURE |
| AI | Irrigated Agriculture |
| AR | Rainfed Agriculture |
| NATURAL VEGETATION |
| NF | Natural Forest and Reforestations |
| NV | Rangeland |
| BARE AREAS |
| BC | Bare Soil Consolidated |
| BU | Bare Soil Unconsolidated |
| BSn | Loose and Shifting Sand |
| BW | Bare Soil in Wadi Environment |
| SABKHAS |
| SK | Terrestrial and Aquatic Sabkha Environment and Waterbodies |
| URBAN AREAS |
| UB | Urban areas, Quarries and Dump Sites |
Mapping Accuracy Assessment
The land cover database developed for Libya was validated through an
accuracy assessment analysis which consisted of undertaking field surveys across
the whole country. This was fundamental to ensure the accuracy of the
photo-interpretation process. Five field work campaigns were undertaken
to carry out in situ validation of 253 randomly selected polygons
representing all the LC features. Each point was reached
using GPS technology. The survey consisted of:
- a general observation of the area around the field point.
- a description of the point itself by completing a specifically modified
Field Form already adopted by the FAO Africover Project, which allows
to verification that the basic interpretation of the point is correct.
The field "Site Description/Observations" was also compiled, when needed,
either to better explain the situation observed or to write down
additional information gathered on the point.
- the collection of at least four pictures of the area surrounding the
point surveyed from different directions, in order to provide a visual
description.
The sum of points collected during the different field trips for the
verification, the accuracy assessment and the change assessment,
generated a database consisting on 5,288 points. The results of the
field surveys confirmed the quality of the land cover database
developed.
Land Cover Change Assessment
A comparative change detection analysis with 1970's, 1980's and
1990-2000's satellite data (Landsat Thematic Mapper) of four sample
areas was also performed. [Details..]
|
|
|
