Estimation of Tree Cover in an Agricultural Parkland of Senegal Using Rule-Based Regression Tree Modeling

Abstract

Abstract:Field trees are an integral part of the farmed parkland landscape in West Africa and provide multiple benefits to the local environment and livelihoods. While field trees have received increasing interest in the context of strengthening resilience to climate variability and change, the actual extent of farmed parkland and spatial patterns of tree cover are largely unknown. We used the rule-based predictive modeling tool Cubist® to estimate field tree cover in the west-central agricultural region of Senegal. A collection of rules and associated multiple linear regression models was constructed from (1) a reference dataset of percent tree cover derived from very high spatial resolution data (2 m Orbview) as the dependent variable, and (2) ten years of 10-day 250 m Moderate Resolution Imaging Spectrometer (MODIS) Normalized Difference Vegetation Index (NDVI) composites and derived phenological metrics as independent variables. Correlation coefficients between modeled and reference percent tree cover of 0.88 and 0.77 were achieved for training and validation data respectively, with absolute mean errors of 1.07 and 1.03 percent tree cover. The resulting map shows a west-east gradient from high tree cover in the peri-urban areas of horticulture and arboriculture to low tree cover in the more sparsely populated eastern part of the study area. A comparison of current (2000s) tree cover along this gradient with historic cover as seen on Corona images reveals dynamics of change but also areas of remarkable stability of field tree cover since 1968. The proposed modeling approach can help to identify locations of high and low tree cover in dryland environments and guide ground studies and management interventions aimed at promoting the integration of field trees in agricultural systems.

Woody plant cover estimation in drylands from Earth Observation based seasonal metrics

“…The same is true for the Landsat based map developed by Sexton et al (2013) (not shown for brevity). This supports existing studies (Herrmann et al, 2013;Gessner et al, 2013) which detect a generally very low tree cover in global products over the Sahel. It further demonstrates that, although the spatial pattern in the MOD44B product is similar and the relation linear to ground data, MOD44B delivers only partial information on the total woody cover in semi-arid drylands, omitting small trees, shrubs and bushes.…”

“…Timesat has been widely used in the Sahel area to detect seasonal dynamics related to the phenology of vegetation and describing the shape of the annual vegetation curve reflected in satellite data (e.g. Horion et al, 2014;Herrmann et al, 2013;Fensholt et al, 2013). The derived seasonal metrics are summarized in Table 1 and illustrated in Fig.…”

“…However, the reliability of these products in semi-arid regions with open tree cover is contested (e.g. Gessner, Machwitz, Conrad, & Dech, 2013;Hansen, Townshend, DeFries, & Carroll, 2005;Herrmann et al, 2013) and limited evaluations against ground observations have been done for drylands in general and for the Sahel in particular.…”

“…Herrmann, Wickhorst, & Marsh, 2013;Karlson, Reese, & Ostwald, 2014;Rasmussen et al, 2011;Sterling & Orr, 2014;San Emeterio & Mering, 2012). However, imageries with a spatial resolution of 1-5 m are cumbersome to process, expensive, susceptible to clouds, and do only provide a static situation for a limited spatial area.…”

Farmlands Feature Greater Woody Cover than Savannas in Semi-Arid West Africa

“…The technical framework of this study adapts local-scale approaches of modeling dryland woody cover 37,38 into reproducible regional/global scale assessments, as the unprece- of woody cover at selected sites served as an independent validation of the remote sensing mapping approach. To achieve a woody cover map of the entire area, the spatially detailed woody cover data derived from VHR images were used to train a gradient boost decision tree regressor to predict woody cover from PROBA-V NDVI and PALSAR-2 images at high resolution (100 m).…”