Repository logo

Geography

Permanent URI for this communityhttps://hdl.handle.net/10413/6543

Browse

Browsing Geography by SDG "SDG15"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Characterization of rainfall over the Limpopo province, South Africa, for the period 1990 to 2020.
    (2024) Moumakwe, Peter Lesiba.; Dube, Lawrence Thembokwakhe.
    The Limpopo Province is home to a large rural population which is highly dependent on rainfed agriculture. Although extensive research has been undertaken to understand rainfall variability over South Africa, a better understanding of the localized rainfall characteristics and variability remains crucial for decision-makers and the livelihoods of the local community. The study aims to investigate rainfall variability and trends over the Limpopo province, understand the distribution (both spatial and temporal) of seasonal rainfall characteristics, and also establish the relationship between seasonal rainfall characteristics with larger modes of climate variability (i.e. El Niño Southern Oscillation (ENSO), and Southern Indian Ocean Dipole (SIOD)). High-resolution Climate Hazard Group Infrared Precipitation with Station Data (CHIRPS) 0.05° gridded data spanning the duration 1990-2020 was employed to analyze the spatial distribution of rainfall over the Limpopo province. In this study dry spells (pentads with < 5 mm rainfall), moderate (rainfall ranging from 10-30 mm per day) and heavy wet days (rainfall > 30 mm per day) were analyzed. Standardized Anomaly Index (SAI) was used to understand the relationship between rainfall characteristics and anomalies. The Mann-Kendall test was also used to determine the trends of seasonal rainfall characteristics over the province. The Pearson correlation was used to establish the association between seasonal rainfall characteristics (dry spells, moderate and heavy wet days) and large modes of variability (ENSO and SIOD). The results of this study show that seasonal rainfall exhibits high spatial and temporal variability over the study period. Throughout the extended summer season (October-March (ONDJFM)), dry spells migrate from the north of Vhembe and Capricorn to the northeast of Mopani, with their frequency and extent increasing from early summer (October-November) to late summer (February-March). The distribution of these rainfall characteristics follows that of mean annual rainfall. Of all periods, December-January (DJ) receives the highest frequency of moderate wet days with a larger spatial extend ranging from 6-13 days in the high-lying escarpment of Vhembe, west of Mopani, south-east of Capricorn, Waterberg, and Greater Sekhukhune. The highest heavy wet day frequency is also observed in the DJ period, over the high-lying escarpment of Vhembe, west of Mopani, south-east Capricorn and north of Greater Sekhukhune records heavy wet days ranging from 3-7 days. The results of the Mann-Kendall trend test revealed a statistically significant decreasing trend in dry spells during DJ and February-March (FM) over the entire Limpopo province. Statistically significant moderate wet day trends were observed during ON over north and east of Mopani, south-east of Capricorn, and west of Mopani district, whereas during the DJ periods, statistically significant increasing trends are recorded over the south-east of Vhembe and northwest of Mopani. During DJ, statistically significant increasing heavy wet day trends are observed over Vhembe, Greater Sekhukhune, and west of Waterberg. The relationship between seasonal rainfall characteristics and rainfall anomalies was observed. The results show that the inter-annual variability of seasonal rainfall characteristics does not always reflect in seasonal rainfall totals/anomalies. This shows that anomalies overlook the isolated impact of seasonal rainfall characteristics. The relationship between seasonal rainfall characteristics and large modes of variability was observed. A strong negative correlation with moderate wet days over the high-lying escarpment in the Vhembe district and south of the Mopani district. However, a complex relationship was observed between the inter-annual rainfall characteristics and large modes of variability. The results showed that not all La Nina years or positive phase SIOD phase equate to wet seasons. Furthermore, years with neutral ENSO and SIOD phases still exhibited above-average wet days.
  • Thumbnail Image
    Item
    Mapping the spatial variability of foliar C:N ratio in a communal rangeland using remote sensing = Ukuhlela ukuguquka kwendawo eyisilinganiso sefoliar C:N ezimfundeni eziyihlanze zomphakathi kusetshenziswa inzwa yokuqapha izimpawubunjalo zendawo.
    (2024) Arogoundade, Mariama Adeola.; Mutanga, Onisimo.; Odindi, John Odhiambo.
    Rangelands contribute significantly to livelihoods by providing grazing land, as well as an array of ecological goods and services. However, they are increasingly threatened by among others, alien invasive plant species, climatic variability and injudicious land management. Hence, sustainable use and optimization of rangelands has recently gained attention. Forage nutrients, such as the C:N ratio are valuable indicators of rangeland quality and quantity, and influence rangeland’s carrying capacity and grazing distribution. Therefore, understanding the spatial distribution of foliar C:N ratio in rangelands is valuable for implementing strategic grazing plans and management strategies. Recently, remotely sensed data, specifically the readily available multispectral sensors with improved spectral properties have gained popularity in foliar nutrients modelling. Consequently, this study sought to model fine scale foliar C:N ratio in a heterogeneous communal rangeland using the new generation multispectral sensors. Thus, five objectives were established, firstly; a review of remote sensing applications in mapping foliar nutrients in tropical grasslands. The findings show that the monitoring of foliar nutrients in grasslands, particularly in Sub- Sahara Africa, using high spatial resolution sensors has been hindered by prohibitive costs. Hence, readily available multispectral sensors remain the most viable option in mapping forage nutrients in heterogeneous landscapes. Secondly; to leverage on Google Earth Engine cloud computing platform to monitor the foliar C:N ratio in a heterogeneous landscape using Sentinel 2 data and the random forest algorithm. The results show an estimated R2 accuracy of 74, with RMSE of 2.68 for the validation datasets of the C:N ratio model established by integrating the spectral bands and vegetation indices. Thirdly, the study sought to test the efficacy of fusing Sentinel 2 and Superdove Planetscope datasets in enhancing the rangeland foliar C:N ratio prediction at a landscape scale. The results demonstrate that freely available new generation multispectral sensors with unique spectral settings offer new opportunities for improving forage C:N ratio mapping in resource-poor countries. Using Sentinel 2 data, the study established that the visible, red edge and near infrared regions of the electromagnetic spectrum were influential in predicting the foliar C:N ratio. The study also established that fusing the spatial resolution of Planet scope with the Sentinel 2’s spectral properties enhanced foliar C:N ratio estimation within a heterogeneous landscape (R2 of 0.79 and RMSE of 2.36).Furthermore, the study noted that both Planetscope's high spatial resolution and Sentinel 2 MSI's high spectral resolutions were valuable in determining the spatial variability of foliar C:N ratio and the inclusion of the red edge spectral settings, combining fused datasets with ancillary variables and the adoption of robust algorithms such as Random Forest improved foliar C:N ratio modelling accuracy. Other variables such as wind effect, topographic wetness index, and the sky view factor also influence the foliar C:N ratio spatial variability . Overall, the findings of this study offer new insights on reliable and cost-efficient approaches for mapping forage nutrients in resource-constrained regions such as South Africa. Using freely available advanced multispectral sensors, the study provides valuable information necessary for optimal rangeland management.