Estimations of Leaf Area Index (LAI) have recently gained attention due to the sensitivity to the effects of climate change and its impact on forest ecosystems. Hence, a study was conducted on the LAI estimation of four vegetation types: (i) gallery forests, (ii) woodland savannas, (iii) tree savannas, and (iv) shrub savannas, at two protected areas of Nazinga Game Ranch and Bontioli Nature Reserve, Burkina Faso. A relationship between LAI and Crown Diameter was also investigated at these two sites. Digital hemispherical photography was used for the LAI estimation. Crown diameters (CD) were determined perpendicular to each other and averaged for each tree and shrub. Overall results revealed that LAI ranged from 0-1.33 and the CD was recorded in the range of 0.46-11.01 m. The gallery forests recorded the highest mean LAI 1.33 ± 0.32 as well as the highest mean CD 7.69 ± 1.90 m. The LAI for the vegetation types were at their lower ends as the study was conducted in summer season, higher values are therefore expected in the wet season, as a significant correlation between LAI and precipitation has been emphasized by various studies. Continuous LAI monitoring and studies on various growth parameters of different vegetation types at the study sites are recommended towards enhanced monitoring and an ecologically feasible forest- and savanna-use and management to maintain essential ecosystem functions and services.
Occupation coding, an important task in official statistics, refers to coding a respondent’s text answer into one of many hundreds of occupation codes. To date, occupation coding is still at least partially conducted manually, at great expense. We propose three methods for automatic coding: combining separate models for the detailed occupation codes and for aggregate occupation codes, a hybrid method that combines a duplicate-based approach with a statistical learning algorithm, and a modified nearest neighbor approach. Using data from the German General Social Survey (ALLBUS), we show that the proposed methods improve on both the coding accuracy of the underlying statistical learning algorithm and the coding accuracy of duplicates where duplicates exist. Further, we find defining duplicates based on ngram variables (a concept from text mining) is preferable to one based on exact string matches.