This study uses a case study approach to adequately validate the readily available GLASOD map [13] in a well studied catchment, where sediment sources have been mapped using field based data and quantitative methods. The experiment was performed in the catchment in which, the author has personally been involved in various initiatives including training, research, and consultancy services. Therefore, it is a case study site on which the author is well acclimatized and knowledgeable, and where sediment flow data are readily available.

The case study area, Nyumba Ya Mungu reservoir catchment, is located in the upstream of Pangani River Basin (PRB), in the North-eastern part of Tanzania and covers an area of about 12,000 km² [10] (Figure 1(a)). It is located between Latitudes 3˚00'00'' and 4˚3'50'' South, and Longitudes 36˚20'00'' and 38˚00'00'' East. The two main tributaries, the Kikuletwa (1DD1) and the Ruvu (1DC1) (Figure 1(a)), join at Nyumba Ya Mungu (NYM), a reservoir of about 140 km² area coverage.

The main sub-catchments in the study area are Weruweru, Kikafu, Sanya, Upper Kikuletwa, Rau, Mue, Himo, Lake Jipe, and Mount Meru slopes. This area has an average annual rainfall of about 1000 mm. The rainfall pattern is bimodal with two distinct rainy seasons, the main rainy season from March to June and the shorter rainy season from October to December. The altitude in the study area ranges between 700 and 5825 m.a.s.l. with Mount Killimanjaro peak as the highest ground. However, the lowlands terrain dominates with coverage of about 73% [10] . Based on the Soil Atlas of Tanzania, the main soil type in the study area is clay with good drainage (Figure 1(b)). Actively induced vegetation, forest, bushland and thickets with some alpine desert chiefly characterize the land cover of the catchment.

Multi-approaches were adopted to identify the sediment sources and erosion processes. The methods herein are: analyses of single hydrological events as sampled from continuous sediment pumping sampler and water levels recording data logger; fingerprinting-organic matter contents and particle size distribution of the transported sediment by rivers or those deposited in the downstream reservoirs (infer the origin and processes of sediment in the catchment); mapping of hydrological variables-rainfall in spatial and temporal domain correlated to sediment transport characteristics at the outlet of the catchment; and numerical modelling. The methods are explained in detail in sections 2.3.1 through 2.3.4 below.

Despite the fact that it was intended to validate GLASOD map with previous findings of the quantitative approach, in addition the map performance was verified by location-based sediment yield rates. It should be noted that the latter are limited in terms of coverage and details. A validated Pacific Southwest Inter-Agency Committee (PSIAC) model [20] was used to estimate sediment yield based on secondary data and field observations [11] . The PSIAC approach is based on a sediment yield classification scheme employing individual drainage basin characteristics (surface geology, soils, climate, runoff, topography, ground cover, land use, upland erosion, channel erosion, and sediment transport). The PSIAC model was built from readily available environmental variables sourced from Tanzania Government's ministries/agencies and public domain global spatial data. Basic data for PSIAC model factor derivation were obtained from topographic maps, geological, soil, land use, ground cover, runoff, climate (mean annual rainfall), and Normalized Difference Vegetation Index map (NDVI). The data were used to generate spatial data layers and to evaluate the sediment factors based on PSIAC concept for the sediment model determination under GIS environment. Each river characteristic was scaled based on PSIAC sediment yield factor rating sheet. All factors characterized by PSIAC model approach were described in a way of acquiring the PSIAC-Indices for each catchment. The PSIAC-Indices for the 31 dams’ siltation data were obtained through preparation, classification and assignment of weights according to PSIAC model building procedures [11] .

Computed Sediment yields for surveyed locations geographically matching with GLASOD map features within the study area were analyzed and reclassified percentiles to represent four (4) severity levels. The percentiles are scaled as 0% - 25% for severity of 1; 25% - 50% for severity of 2; 50% - 75% for severity of 3; and 75% - 100% for severity of 4. Field data and GLASOD map severities were correlated. A Student’s t-distribu- tion table was used to confirm the strength of correlation. The correlation was considered strong if a computed t value is greater than table value at 5% level of significance as recommended by [21] . Besides, coefficient of determination, R², values greater than 0.5 were considered acceptable as suggested by other researchers [22] .

In this section of the paper a detailed explanation on the identified sediment sources and erosion processes based on extracted attributes from GLASOD map is provided. For this purpose, Figure 2 and Table 3 and Table 4 are

referred to Figure 2, a generated GLASOD map as described in section 2.2.2 presents degradation severity classes for all 20 subcatchments of the Nyumba Ya Mungu Reservoir catchment. You would see that there were 4 severity classes mapped in the study area. This suggests that the study area is heterogeneous, that is to say soil degradation phenomena is spatially variable. However, one could see that there is a general spatial pattern on soil degradation status. Severity classes for upstream part of mountainous catchments, along Mts. Kilimanjaro and Meru slopes, have lower severity classes than the downstream parts of respective catchments. An exception trend is observed on the eastern part of the study area, especially on the Pare Mountains (upstream of Lake Jipe), where degradation is characterised by higher severity classes. It is noteworthy that severity class of 1 dominates the study area with coverage of about 35 percent (Figure 2; Table 3 & Table 4). Severity class of 4 with a least coverage represents only 15 percent of the area. This suggests that erosion prone areas or sources are localized within the study area.

Table 3 presents the status of soil degradation considering the type, extent, degree, rate and causative factors of degradation within physiographic units, sub-catchments, in the study area. The table shows that, with exception of 8 sub-catchments (i.e., 11, 12, 13, 14, 15, 16, 17, and 19), there are two types of soil degradation recognized in each mapped unit. That is why an aggregated severity (severity class) is presented/provided as well. Where a soil degradation type 2 is nonexistent, a value of zero (0) is assigned to soil degradation characteristics. For map units with two causative factors, the sequence of appearance in the last column of Table 3 does neither indicate a sequence in importance, nor it necessarily coincide with the sequence of degradation types indicated in the table. As depicted in the table the aggregated severity is sometimes higher than severity of the individual degradation type. For some mapping units where one of the two degradation types is subordinate, the aggregated severity is one class higher than the severity of the most important type. This always occurs when the severity of the second type is significant enough to have a bearing on the overall severity [5] . The degree of degradation, relative extent of degradation and causative factors are used in combination to define the severity of soil degradation. For instance, severity code of Wt2.3.g/#, in Table 3, sub-catchment number 1, the two letter codes (Wt) identify the type of soil degradation. This letter combination is followed by two numbers: the first number (2) refers to the degree; the second number (3) refers to the relative extent of soil degradation with overgrazing (g) as the causative factor and the severity class upgraded (#). Wt2.3 therefore means that the degradation type “loss of topsoil through water erosion” (Wt) has a moderate degree (2) and occurs frequently (3) caused by overgrazing with severity class (#) upgraded.

For clarity purpose, the characteristics in Table 3 are summarized and expounded in Table 4. In Table 4 soil degradation characteristics are presented for all 20 sub-catchments in the study area. Such presentation format will be useful in comparing between GLASOD and previous studies’ results. For such purpose, other attributes including catchment name and area coverage are introduced. Table 4 indicates that there are two main degradation types, viz., loss of top soil through water erosion and terrain definition/mass movement through water erosion with degree of degradation varying from light to strong. With the exception of Lake Jipe sub-catchment, the downstream parts of the sub-catchment soil degradation are in the form of mass movement. It should be understood that the most common phenomena of Terrain definition/mass movement degradation type as defined on GLASOD map are rill, gully formation, riverbank destruction and landslides [5] .

Loss of top soil through water erosion (sheet erosion) degradation type is found in the upstream parts of the catchment. As farming activity in the study area is practiced on the upper sub-catchments, slopes of Mts. Kilimanjaro and Meru, it is perceived that topsoil is rich in nutrients. Therefore, a relatively large amount of nutrients may be lost together with the topsoil. Such degradation type may lead to an impoverishment of the soil. In this context, on very steep slopes of Mts. Kilimanjaro and Meru, natural loss of topsoil may occur frequently. Unfortunately, this “geologic erosion” could not be indicated on the GLASOD map. Degree of degradation in the Lake Jipe sub-catchment is very strong and besides it experiences both sheet erosion and mass movement degradation types. In general the principal external dynamic agent of erosion is the hydrospheric forces of water, i.e., rainfall, runoff, and stream flows. The data from table also suggest that soil degradation is caused mainly through removal of natural vegetation and overgrazing. The latter is practised in major parts of the study area with coverage of 85 percent. The terrain in the overgrazed area could be characterized as low lands and sparsely vegetated. The two causative factors, removal of the natural vegetation and overgrazing are interrelated and have interplay role in triggering erosion. According to [5] , on GLASOD map, removal of the natural vegetation is normally attributed to land reclamation activities such as farming, cattle raising, road construction, and urban development. On the other hand, the effect of overgrazing is linked to livestock trampling. Trampling may cause