Development of provisional guidelines for the treatment of scale and resolution in assessing streamflow reduction impacts of alien plant infestations and commercial afforestation in water resources modelling studies

Dzvukamanja, T. N. (Tembiwe Nunana) (2005-12)

Thesis (MScEng)--University of Stellenbosch, 2005.

Thesis

ENGLISH ABSTRACT: Experiments conducted on afforested catchments in South Africa have shown that alien trees can cause substantial reductions in catchment runoff (Scott et ai, 2000). In recognition of the impact which alien trees can have on the country's water resources, commercial afforestation was declared a stream flow reduction activity (SFRA) in terms of the National Water Act (NW A) (No. 36 of 1998), and the Department of Water affairs and forestry launched the Working for Water Programme (WfW) in 1995 with the recovery of water resources lost to Invasive alien plants (lAPs) as one of the Programme's objectives. These initiatives have intensified the need to quantify SFR; for example, for licensing purposes to satisfy the requirements of the NWA and for predicting the effects of lAP clearing by WfW projects. Of interest to water resources practitioners, is the impact of SFR on mean annual runoff (MAR), on low flows and on water resource system, or reservoir, yield. In South Africa two basic methods of streamflow reduction (SFR) estimation have been developed for commercial afforestation and lAPs. These are • free-standing empirical relationships in the form of the CSIR SFR CUNes, used in conjunction with the monthly, calibration-based, Pitman model. • component modules in the physically-based, land-use sensitive ACRU rainfall-runoff catchment model, run at a daily time step with relatively fine subcatchment delineation. There has been a strong need for an evaluative comparison of the impacts of SFR estimated via these two methods. This study aimed to meet this need by using both methods to estimate SFR for a number of commercial afforestation and lAP scenarios in three study systems, the Berg, Sabie and Mhlatuze, representing different bioclimatic conditions in South Africa, and running the SFR sequences from the two estimation methods through the Water Resources Yield Model to determine the impact of the SFR on yield. The analysis differentiated between upland and riparian SFR, and between SFR produced by different tree classes. Study conclusions included the following points: • Both the ACRU and SHELL models are capable of achieving a reasonable average seasonal correspondence of high and low flows with the observed averages, though the actual averages produced by the two models can differ substantially. In general, ACRU simulates less SFR than SHELL, and gains in SFR after afforestation or invasion by lAPs may be simulated by ACRU during dry periods. The selection of crop factors for different plant species has a strong influence on the relative water use of the species modelled in ACRU. • The impacts on yield of SFR due to lAPS and afforestation tends to be greater than the impact on MAR, and impacts tend to be more severe for small subcatchments than for the total catchment. A simulated reduction in MAR can result in a simulated increase in yield of a given assurance, if the portion of the flow sequence occurring during the critical period is dominated by streamflow gains, and vice versa. Research recommendations centred on improving the availability of reliable field measurements of parameters and processes required tor the effective modelling of SFR. Based on the results of the study, guidelines were formulated for SFR modelling, focussing on the choice of SFR estimation method and the treatment of various parameters and considerations which influence the outcomes of SFR modelling.

AFRIKAANSE OPSOMMING: Eksperimente wat in bebosde opvanggebiede in Suid-Afrika uitgevoer is, het getoon dat uitheemse bome aansienlike verminderings in opvanggebied-afloop kan veroorsaak (Scott et al,2000). Ter erkenning van die impak wat uitheemse bome op die land se waterbronne kan he, is kornrnersiele bebossing verklaar as 'n stroomvloei-verminderingsaktiwiteit (SVVA) in terme van die Nasionale Waterwet (NWW) (Nr. 36 van 1998). Die Departement van Waterwese en Bosbou het ook die Werk-vir-Water Program (WvW) in 1995 geloods met, as een van die doelwitte, die herwinning van waterbronne wat deur uitheemse indringerplante (UIPe) opgebruik word. Hierdie inisiatiewe het die behoefte om SVV te kan kwantifiseer verskerp; by voorbeeld; for liksensiermqsdoeleindes om die vereistes van die NWW te bevredig, of om die impakte van UIP-opruiming in WvW-projekte te voorspel. Van besondere belang vir waterbron-praktisyns is die impak van SVV op gemiddelde jaarliks afloop (GJA), op lae vloeie en op die lewering van waterbronne, of -stelsels. In Suid-Afrika is twee basiese metodes vir SVV-raming ontwikkel vir komrnersiele bebossing en UIPe, soos volg: • losstaande empiriese verbande in die vorm van die WNNR se SVV-krommes, wat gebruik word saam met die maandelikse, kalibrasie-gebaseerde, Pitman-model wat in die SHELL-sagteware-omgewing ingebou is. • Modules wat komponente vorm in die fisies-gebasserde, grondgebruik-gevoelige ACRU reenval-atloop opvanggebiedmodel, wat op 'n daaglikse tydstap loop, met relatiewe fyn subopvanggebied-indelings. Daar bestaan al lank 'n sterk behoefte aan 'n takserende vergelyking van die impakte van SVV so os geraam via hierdie twee metodes. Hierdie navorsing het beoog om hierdie behoefte te bevredig deur beide metodes in 'n aantal kornrnersiele bebossings- en UIPscenario's in drie stelsels, die Berg, Sabie en Mhlatuze, te gebruik. Sodoende word drie verskillende bio-klimaatstreke gedek. Die maandelikse SVV-tydreekse van die twee ramingsmetodes was toe ingevoer in 'n waterbronstelselmodel (WRYM) om die impak van die SVV op die lewering te bepaal. Die ontledings het tussen oewer- en nie-oewer-SVV, asook tussen SVV wat deur verskillende boom-klasse veroorsaak is, onderskei. Die gevolgtrekkings uit die studie het die volgende punte ingesluit: • Seide die ACRU- en SHELL-modelle is in staat om 'n redelike ooreenkoms in seisoenale hoe en lae vloeie met waargenome gemiddeldes te verskaf, alhoewel die eintlike gemiddeldes wat deur die twee modelle gelewer word, aansienlik kan verskil. • Oor die algemeen simuleer ACRU laer SVV as SHELL en klein toenames in vloeie na bebossing of indringing deur UIPe kan soms tydens droe tydperke deur ACRU gesimuleer word. Die keuse van gewasfaktore vir verskillende planttipes het 'n groot invloed op die relatiewe waterverbruik van die planttipes wat in ACRU gemodelleer word. • Die impakte op lewering van SVV te wyte aan beide UIPe en bebossing neig om groter te wees as die impak op GJA, en die impakte neig om meer ernstig te wees vir klein subopvanggebiede as vir die totale opvanggebied. 'n Gesimuleerde vermindering in GJA kan soms saamval met 'n gesimuleerde toename in lewering teen 'n spesifieke betroubaarheid, as die gedeelte van die tydreeks wat gedurende die kritieke tydperk voorkom, heelwat UIP-gebaseerde stroomvloei-toenames bevat en vice versa. Navorsingsaanbevelings fokus op die verbetering van die beskikbaarheid van betroubare veldwaarnemings van parameters en prosesse wat vereis word vir betroubare modellering van SVV. Riglyne vir SVV-modellering is geformuleer, gebaseer op die resultate van hierdie navorsing, met 'n fokus op die keuse van SVV-ramingsmetode, die behandeling van verskeie parameters en oorwegings wat die uitslag van SVV-modellering sou kon bemvloed.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/50319
This item appears in the following collections: