Hydro-engineering solution for a sustainable groundwater management at a cross border region: case of Lake Nyasa/Malawi basin, Tanzania
- First Online:
- 581 Downloads
A groundwater resources assessment has been carried out for the Lake Nyasa Basin east Africa, with reference to sub catchments further to the whole basin wise analysis, including quantification of potential yields from the aquifers. Numerical groundwater models, MODFLOW-SURFACT coupled with Visual MODFLOW, were used to assess sub-basin groundwater sustainability (Via “Zone Budget” software package). The model has been calibrated to observed hydraulic head and processed baseflow estimate. It is concluded that the aquifer system is sustained by episodic recharge and the long-term gaining storage which represents the maximum extractable volume. Future predictions of groundwater usage indicate that by 2035 the percentage of annual safe yield extracted will increase to between 11.3 and 103%. Model result suggests that there is a need to revise the current estimate of sustainable yield based on future climate change conditions and projected population growth, which would decrease spring flows substantially and decrease hydraulic head basin-wide. It also suggests that by 2035 some sub catchments will be nearly at or exceeding the annual safe yield leaving no room for socio-economic development, or the need to reduce existing socio-economic demands to meet domestic demands. Besides to improving the natural replenishment capacity through artificial recharge technique, the other option is to increase the percentage of total recharge allocated to the annual safe yield from 10% of the total annual recharge to 20% in these catchments. This provides a basis for management of individual groundwater scheme in sustaining livelihood activities or implications of policies and to develop a plan for potential groundwater extraction scenarios pertaining to water use and allocation.
KeywordsWater resources management Lake Malawi/Nyasa Tanzania East Africa Rift valley
Groundwater models are valuable tools for the catchment managers to take actions on the management of groundwater resources [1, 2, 3, 4]. Groundwater occurrence and movement are controlled primarily by the aquifer permeability and the lithology of the underlying strata [5, 6]. In groundwater-resource assessments, terms such as sustainability, sustainable yield, sustainable pumping, sustainable development, and safe yield are often used interchangeably or with little distinction between the concepts. Definitions and approaches to assess sustainability of potable drinking-water resources are summarized by Bredehoeft , Alley and Leake  and Devlin and Sophocleous  among others. Sophocleous , for instance defines sustainable development as development that meets the needs of the present without compromising the ability of future generations to meet their own needs. Specifically, optimal or sustainable groundwater-resource use requires setting upper limits on water withdrawal (or sustainable yield) to avoid compromising the source . Previous researchers (e.g. ) defined safe yield as the maximum quantity of water which can be extracted from an underground reservoir, yet still maintain the supply unimpaired. Under natural conditions, recharge is balanced by discharge from the aquifers by evapotranspiration and/or exfiltration into streams, springs, and seeps . It is widely recognized that there is a need to reserve a fraction of the recharge for the benefit of surface waters, related ecosystems and that fraction is difficult to quantify [13, 14, 15, 16].
In areas where potable water resources are limited, an accurate and reliable estimate of sustainable use or yield is critical. Yield estimates have to address not only decreasing water availability, as manifested by declining water levels, but also deterioration of water quality caused by, e.g., seawater intrusion or agrochemical use. In recent years, hydrological models with various degrees of sophistication, have been used in different settings to evaluate water budgets under several climate or land-use scenarios. Several approaches have been suggested for evaluation of sustainable yield . Generally, sustainable yield of small-scale abstraction schemes is estimated on the basis of analytical solutions applied to constant-rate test data and forward modelling with the estimated transmissivity and storativity. The modelled abstraction rate is adjusted to limit the drawdown to a predefined level (usually the uppermost water strike) at the end of a designed production period. Difficulties arise in applying analytical solutions where the simplifying assumptions are not supported by field data . Numerical models are used in hydrogeology to increase understanding of the complex systems that characterise aquifers [19, 20]. The results obtained from these models cannot be considered complete or totally accurate because of the inherent uncertainty of the models [21, 22]. However, the main trends of the hydrological behaviour may be identified, as long as the underlying conceptual model is a reasonable reflection of reality, and the parameters of the model have physically reasonable values [7, 23, 24]. Once these conditions have been met, the numerical models may be used as decision-making tools in water-management planning .
Lake Nyasa Basin, situated in the south western part of Tanzania is part of the greater Zambezi Basin. Lake Nyasa is an African Great Lake and the southernmost lake in the East African Rift system. The lake is 560–580 km long, has a maximum width of 75 km and an average depth of 292 m and a maximum depth of 706 m. It drains via the Shire River in Mozambique which flow to the Zambezi River which discharges to the Indian Ocean. Lake Nyasa is the ninth largest and third deepest, fresh water lake on earth  with an estimated average volume of around 8000 km3. The lake supports fisheries, livestock, agriculture, tourism and wildlife in all three countries [26, 31]. Lake Nyasa and the associated catchment basin is shared between Tanzania, Malawi and Mozambique and covers a total area of 165,109 km2 including the lake. The lake catchment within Tanzania is calculated to be approximately 27,500 km2. For this study, the portion of the basin which falls within Tanzania, will be referred to as the “basin”. The basin within Tanzania has a large topographic range, being from approximately 470 masl at Lake Nyasa to 2960 masl at Mount Rungwe in the northern part of the basin. The elevation of Lake Nyasa varies annually by around 2 m depending on rainfall .
Compilation of drilling completion, chemical analysis, climate data, soils, geology, topography, vegetation, landuse, soils and population information maps and existing reports were carried out for assessment of the geology, geomorphology, stratigraphy of the basin, establishment of groundwater resource inventory, groundwater potential evaluation, determine productivity for the basin and groundwater recharges mechanism using GIS (e.g. [5, 6]). A total of 316 borehole records were located. The deepest average Standing Water Levels (SWL’s) are found in Ruhuhu, Mchuchuma and Kiwira catchments (Fig. 2), for weathered and fractured rock and range from 20 to 28 m below ground level (m bgl). The shallowest average SWL’s are found in Mbaka, Kiwira alluvials, Lufirio, Songwe and Mbawa where the average SWL’s are between 2.2 and 10 m bgl. It is also likely the depth to water is strongly controlled by topography and climate as opposed to geology.
The Kyela flood plain alluvium has the highest average yields at 21 m3/h followed by Mchuchuma with 20.5 m3/h. Average yields for all other areas, which are screened in weathered to fractured rock are low, being less than 8 m3/h. The lowest yields are found in Mwaba with a range of 0.4–2.2 m3/h and an average of 1.26 m3/h.
Aquifer hydraulic parameters
Assessment of the pumping test data was made using Aqtesolv Pro® Version 4.5. The alluvials have a transmissivity and hydraulic conductivity range of 1.16–761 m2/day and 0.01–7.6 m/day respectively. The transmissivity for the weathered and fractured rock have a range of 0.3–17.7 m2/day. The permeability ranges from 0.01 to 1.0 m/day. Although these results indicate that the groundwater potential is very high for the alluvials, the permeability is generally low. On the other hand, while the groundwater potential for the basement rocks is low to moderate, the permeability is generally low to moderate.
Specific capacity was used to identify well or borehole potential, with high values indicating good potential. Typically the water level drawdown was recorded after 24 h of pumping to allow stabilisation of the water level and development of a cone of depression. Shorter time periods are not considered reliable. The specific capacity range of the alluvials and weathered/fractured rock are 129–4089 m3/h/m and 186–11,640 m3/h/m respectively. The specific capacity data indicates that alluvials can have nearly three times the potential of bores in the weathered and fractured rock aquifers . This is in contrast to the yield information, but likely is a better reflection of the actual potential which is related to the thickness of the water bearing zone. This effectively links to aquifer types limiting the usefulness of the data. However assessment of the transmissivity from the data is considered reliable enough for providing a general basin wide estimate of aquifers potential.
75% of Tanzania is underlain by largely Precambrian crystalline basement complex rocks with various composition and ages, which form the basement aquifers. Other aquifer types include sandstone, coastal sedimentary formation of limestone, and alluvial sedimentary sequences, and volcanic materials. The groundwater potential of each type of aquifer differs significantly at the local scale as well as at the basin scale. The occurrence of groundwater within the Basin is most influenced by the geology/geomorphology, followed by topography, rainfall, and all of which influence recharge and discharge from the aquifer systems. However development of groundwater within the basin is most influenced by topography and the associated drainage patterns, with the exception of the Kyela area (Fig. 1). The groundwater potential for the basin is generally considered low but locally moderate with no high yield systems, except the Kyela alluvials. However the lack of borehole data and poor distribution prohibits an accurate assessment of the potential.
Numerical simulation of hydraulic head and processed baseflow
The physical system described in the previous section provides a framework for setting up and calibrating a steady numerical groundwater flow model. The objective of groundwater modelling is to gain an understanding of the groundwater flow system in the Lake Nyasa basin and to help constrain and elucidate groundwater contribution to baseflow. The model is parameterized based on hydrostratigraphic zones and constrained using the aforementioned data (hydraulic head and baseflow). A three dimensional model is used for better reproduction of the complex topography and heterogeneity of the watershed. Head, river flow, precipitation data, and the geology and surface topography of the region are used to construct the model. Model output includes distribution of hydraulic head and water budget targets. Yield is calculated from the model output.
The model was constructed based on background information provided and the assessment of available information in the previous sections. Modelling was undertaken using MODFLOW-SURFACT code , an advanced MODFLOW based code developed by HydroGeoLogic Inc. that handles complete desaturation and resaturation of grid cells, within the framework of Visual MODFLOW Version 4.6. The pseudo-soil function of MODFLOW-SURFACT was chosen to account for the saturated and unsaturated processes within the model, as it avoids the wet/dry non-linear iterations of standard MODFLOW. “Zone Budget” software package were used to assess sub-basin groundwater sustainability.
Spatial discretization and layering
- Layer 1: used to simulate the upper most aquifer including the alluvium, weathered basement and shallow fractured basement (Fig. 6) with a thickness of 5–250 m; and
Layer 2: A non-uniform deep basement layer ranging in thickness from 400 m to 1500 m.
Boundary and initial conditions
Recharge and discharge
Recharge is assumed to be predominantly from rainfall infiltration. This may occur as direct recharge through exposed bedrock fractures or via slow percolation through the weathered overburden. A small component of recharge is thought to occur via stream and river bed leakage especially in the lower reaches of the river systems. As there is no long term water level monitoring data recharge has been assumed to be a percentage of the total annual rainfall . Rainfall infiltration rates in the order of 5–20% have been estimated during the model calibration. Recharge zoning was made based on the geology and rainfall isohyet map.
Groundwater discharge is via evapotranspiration, baseflow of streams and rivers (as river leakage which discharges to the Lake), pumping, and deep percolation discharge to Lake Nyasa. The component of deep percolation is generally assumed to be about 10% of recharge. In the elevated plateaus of the catchments, baseflow is assumed to be the major discharge mechanism. In the lower reaches, evapotranspiration is thought to play a larger role especially in the lower reaches of the Ruhuhu River and Kyela alluvials. Given the depth of groundwater is generally shallow in the model area, the evapotranspiration (ET) is considered significant. The ET rate was set at 10% of the estimated values shown on Fig. 2, with an extinction depth of 2 m. The ET rate was refined during the calibration process.
Adopted aquifer properties layer 1 (for the zone color refer to Fig. 8)
Color of unit
Horizontal hydraulic conductivity (K) (m/day)
Vertical hydraulic conductivity (K) (m/day)
Consolidated sedimentary rocks (Karroo)
Basalt and mafic rocks
Metasediments, undifferentiated, 1 and 2
Light grey and yellow
The initial head was interpolated from limited water levels recorded is used as the starting point for model calibration. The interpolation was used to establish precalibration steady-state conditions.
Volumetric budget for entire model at end of the time step 1 in stress period
Rates for this time step
In − out
In − out
Baseflow measurements in streams and rivers may be used as additional calibration targets as they allow calibrating to a flow rate as a head distribution, which may greatly increase the uniqueness or confidence of the model calibration. For this study, stream flow measurements from two locations on the Kitiwaka River recorded during an extended dry period on the 16th December 1983 were selected for the analysis (indicated by a cross line symbol on Fig. 6). The difference between these flow measurements is the net baseflow to the stream, assuming evaporative losses are negligible.
Results and discussion
Simulated water balances
Model groundwater zone budget results between the stream gauge stations
River leakage (baseflow)
From external zone
A negative result indicates a net loss of water from the aquifer to the Kitiwaka River as all flow rates in MODFLOW are considered with respect to the aquifer. The recorded gain in net flow is 2.05 m3/day while the model indicates a gain of 1.85 m3/day. The difference (0.2 m3/day) is small (~90% similarity) and well within the error of the model. The model result indicates that net loss of water from the aquifer may explain ~88% of the difference in stream flow between the two stream flow gauge stations. This is in agreement with the estimated baseflow component processed using baseflow Index  method (i.e. 85%).
Volumetric budget for each catchment
Volumetric water budget for catchments
Inflow (1000 m3)
Outflow (1000 m3)
Recharge from rainfall
From other catchment
Constant head (Lake Nyasa)
To other catchment
Projected population by catchment
(based on National Bureau of Statistics, Tanzania—2002 census data)
Annual safe yield/extraction limit
Estimated safe yield for each catchment and current groundwater usage
Recharge volume (1000 m3)
10% recharge (annual safe yield) (1000 m3)
Estimated current groundwater usage (1000 m3)
% of ASY currently used
Remaining groundwater available for development (1000 m3)
To assess the potential groundwater resource for the basin the recharge has been estimated using the groundwater model. The estimated sustainable safe yields from the groundwater model for each catchment are compared to the estimated current domestic groundwater usage with a 20% contingency added for stock watering, small scale irrigation, commercial, institution (schools and Government buildings) and losses. Industrial, Mining and large scale agriculture are not included, they are considered socio-economic demands which may be developed after domestic uses are covered. It can be seen (Table 6) that currently the percentage of annual safe yield extracted ranges between 6.8 and 62%, averaging around 29%. This suggests there is room for socio-economic expansion of groundwater usage within all catchments.
The annual safe yield of the basin is taken as 10% of the total annual recharge and calculated to be 10.7 × 107 m3. Current conservative estimates of groundwater use within the basin are around 27.3 × 105 m3 which is around 30% of the basin annual safe yield. The current catchment groundwater usage ranges between 10 and 60% of the annual safe yield leaving between 90 and 40% of the annual safe yield depending on sub catchment available for future development . Based on the current model result, estimate of groundwater recharge for the study area is about 1.07 × 106 m3/day. Thus, the limits on aquifer pumping as a percentage of groundwater recharge range can be computed for the sustainable yield. Elsewhere, a value of total use of 50–70% of the sustainable yield was adopted.
Projected total groundwater usage, % of ASY and remaining available for development
Total GW usage
% ASY used
GW available for other uses
Total GW usage
% ASY used
GW available for other uses
Total GW usage
% ASY used
GW available for other uses
Impacts on groundwater due to the population patterns may include increased abstraction for domestic, industrial or agricultural purposes, decreased recharge through land clearing and urbanisation, contamination from sewage and chemicals and localise increases in recharge due to storm water soakaway’s in urban areas. Population concentrations in areas with limited access to gravity schemes, such as the south eastern part of Ruhuhu catchment near Songea, have resorted to groundwater for domestic water supplies.
These predictions do not account for possible reductions in the total annual recharge as a result of climate change . It can be seen that by 2035 the percentage of annual safe yield extracted will increase to between 11.3 and 103%, averaging around 50%. This indicates that less water may be available for socio-economic development in the future. It also suggests that by 2035 Kiwira and Mbaka sub catchments will be nearly at or exceeding the annual safe yield leaving no room for socio-economic development, or the need to reduce existing socio-economic demands to meet domestic demands. The other option is to increase the percentage of total recharge allocated to the annual safe yield from 10 to 20% in these catchments.
Basin sustainability yield constraints
The current study shows that derivation of sustainable yield using conservation of mass principles leads to expressions for basin sustainable that can be readily determined using numerical modeling methods and selected on the basis of applied constraints. Bores are concentrated in areas of high population. For instance the village of Mlete, in Tanga Ward, Namtumbo District has 21 identified bores. The yield from each bore may not be high, but the combined yield of all 21 bores may be. Viewed on a large scale this is likely not an issue, however the combined yield of the bores may exceed the safe yield in a small area leading to localised drawdown.
The possibilities of severe, long-term droughts and climate change also should be considered. Because a climate stress on the hydrologic system is added to the existing or projected human-derived stress, droughts represent extreme hydrologic conditions that should be evaluated in any long-term management plan. Water reuse is one of the potential ways to complement the existing portable water sources. Practice of the water reuse for irrigation in Tanzania and similar region has been suggested by Kihila et al. ; Yihdego . The challenges that need to be addressed, the benefits expected and global wastewater treatment options, capital and operation costs as well as the cost benefit analysis, implementation of water reuse will require among other things, proper defined policy and institutional framework, clear guidelines as well as more research and investment on wastewater treatment have been pointed out .
When considering the groundwater potential and annual safe yield for the basin the potential for Transboundary impacts must be considered. Lake Nyasa is shared between three countries while the basin borders three countries, Zambia and Malawi along the north western boundary marked by the Songwe River and Mozambique on the southern boundary (Fig. 1). The Songwe River marks the border of Tanzania with Zambia and Malawi, the river is gaining having a high groundwater baseflow. On the Zambia/Malawi side are high hills and mountains which fall towards the river indicating the groundwater flow direction is also towards the river. Extraction of groundwater within the basin (Tanzania) is unlikely to impact groundwater across the border and the potential for transboundary impacts are considered very low. However over extraction of groundwater may impact baseflow to the river and therefore surface flows in the river which may impact users on both sides of the border. The southern basin boundary with Mozambique is an arbitrary line which doesn’t follow topographic features, therefore there is a potential for transboundary impacts. However the length of border is short and in a mountainous area with high rainfall and low population limiting the potential for extraction. The potential for transboundary groundwater impacts are therefore considered low to very low.
To analysis the groundwater development based on sub-basin/sub-catchment, a three dimensional groundwater numerical modelling (MODFLOW-SURFACT coupled with Visual MODFLOW) were used to handle Vadose Zone and for greater numerical convergence stability. Groundwater modelling and water balance optimization were used for the sustainable yield estimation. The model has been calibrated to the hydraulic head and processed baseflow estimate, which adds confidence or none-uniqueness to the solution and hence as a basis good enough for prediction/scenario simulation. A water-budget model output was used in estimating recharge which allows for an evaluation of the future groundwater sustainability, with several sub-basin extraction limits/annual safe yields in the Lake Nyasa basin. The water balance error was 0% which could be considered as criteria (i.e. water balance stabilization), to constrain the estimate of sustainable extraction . Based on the current model result, estimate of annual safe yield based on 10% of groundwater recharge for the study area is about 10.7 × 107 m3. The current catchment groundwater usage ranges between 10 and 60% of the annual safe yield leaving between 90 and 40% of the annual safe yield, depending on sub catchment, available for future development. Future predictions of groundwater usage indicate that by 2035 the percentage of annual safe yield extracted will increase to between 11.3 and 103%. Sustainability indicators such as acceptable head can be examined based on climate and management decisions affecting water extraction and land use. The choice of the method employed in this study has proved successful for a sound understanding of the sub-basin groundwater resources sustainability among the stake holders, decision makers for an effective management. With this information, society can make better informed decisions about how to manage their ground water resources in a long-term context. Such analyses also ideally lead to the design and implementation of long-term hydrologic networks to monitor projected outcomes of the ground water development and to improve the ability to predict future system responses.
The basin size is relatively small and annual groundwater recharge volumes are low, assuming between 5 and 20% of rainfall and therefore sustainable groundwater extraction volumes will be low, and there is insufficient information to assess the component of deep circulation and hence the deep groundwater resource within the fractured rock systems. The possibility of transboundary issues are considered low due to the prevailing groundwater gradient in the areas of international borders. It is more likely groundwater extraction may lead to reduced baseflows especially in the Songwe River which may impact surface water users on both sides of the border. Basin groundwater quality is generally good due to the annual recharge of the shallow aquifers. Based on the limited available groundwater analysis all aquifer types in all catchments may be suitable for domestic and stock watering purposes subject to assessment of individual bores.
The effect of few management options and a simplified climatic-change scenario needs to be considered for the Lake Nyasa sub-basin aquifers together with a decrease in the net recharge (and possibly suggest Managed Aquifer Recharge/Aquifer Storage recovery), a change in river–aquifer interactions, and the total abstraction volumes identified as the main factors to address in management planning. Further efforts to constrain the hydrogeological parameters of the various geological units, including riverbed deposits, which could include targeted test pumping, could improve understanding of the system.
YY performed numerical simulation and wrote most of the sections of the paper. AP analysed the field data and wrote some sections of the paper. All authors read and approved the final manuscript.
We acknowledge SMEC International and Tanzanian Ministry of Water. The study was funded by the International Development Association and the United Republic of Tanzania.
We would like to extend our thanks to Lake Nyasa Basin Water Board; Sokoine University of Agriculture, Tanzania; Drilling and Dams Construction Agency (DDCA), Dar es Salaam Office; Songea District Water Engineers Office, and Kyela District Water Engineers Office. The manuscript has benefitted from the reviewers’ and editors’ comments.
The authors declare that they have no competing interests.
- 13.Smith A, Walker G, Turner J (2010) Aquifer sustainability factor: a review of previous estimates. International Association of Hydrogeologists (AIH) and the Geological Society of Australia (GSA). EP104589Google Scholar
- 26.Danish International Development Agency (DANIDA) (1988) Implementation of water master plan for Iringa, Ruvuma and Mbaya regions, for United Republic of TanzaniaGoogle Scholar
- 27.Don Consult Ltd (2011) Provision of technical and facilitation consultancy services for rural water supply and sanitation sub-projects in Kyela district council, exploration drilling report. Tender No: LW/DC/F. 10/15/Vol. V/108, July 2008. Report for the United Republic of Tanzania, Prime Minister’s Office Regional administration and Local Government, Water Sector Development ProgramGoogle Scholar
- 29.SMEC (2013) Integrated water resources management and development plan—groundwater assessment. Ministry of Water, Dar es SalaamGoogle Scholar
- 30.Bootsma HA, Jorgensen SE (2006) Lake Malawi/Nyasa. Experience and lessons learned brief. http://Worldlakes.org. Accessed 25 March 2015
- 31.Kashaigili JJ (2010) Assessment of groundwater availability and its current and potential use and impacts in Tanzania, Report for International Water Management Institute, December 2010Google Scholar
- 32.Lake Nyasa Basin Water Board (LNBWB) (2010) Basin (Lake Nyasa) hydrological report, December 2010Google Scholar
- 35.Smakhtin VU (2001) Estimating continuous monthly baseflow time series and their possible applications in the context of the ecological reserve. Water SA 27(2):213–218 (ISSN 0378–4738) Google Scholar
- 36.Howard AE (2011) A brief introduction to the geology and mining history of Tanzania. Report prepared for SIKA RESOURCES INC. Nebu Consulting LLC, TorontoGoogle Scholar
- 38.HydroGeoLogic Inc (2002) MODFLOW-SURFACT: a comprehensive MODFLOW-based hydrologic modelling system: code documentation and user’s manual. HydroGeoLogic Inc, RestonGoogle Scholar
- 40.Fetter CW (2001) Applied hydrogeology, 4th edn. Prentice Hall, Englewood CliffsGoogle Scholar
- 41.Barnett B, Townley LR, Post V, Evans RE, Hunt RJ, Peeters L, Richardson S, Werner AD, Knapton A, Boronkay A (2012) Australian groundwater modelling guidelines, waterlines report. National Water Commission, CanberraGoogle Scholar
- 43.Middlemis H (2004) Benchmarking best practice for groundwater flow modelling. The Winston Churchill Memorial Trust of AustraliaGoogle Scholar
- 44.Harbaugh AW (1990) Zone budget. A computer program for calculating subregional water budgets using results from the US geological survey modular three-dimensional ground-water flow model: US geological survey open-file report 90–392Google Scholar
- 47.Fornés JM, De La Hera Á, Llamas MR (2005) The silent revolution in groundwater intensive use and its influence in Spain. Water Policy 7:253–268Google Scholar
- 49.Yihdego Y (2015) Water reuse and recreational waters, urban water reuse handbook (UWRH), chapter 78. In: Eslamian S. (ed) Taylor and Francis, CRC Press, Boca Raton, pp 1029-1039. Print ISBN: 978-1-4822-2914-1. eBook ISBN: 978-1-4822-2915-8. CAT# K22608. https://www.crcpress.com/Urban-Water-Reuse-Handbook/Eslamian/p/book/9781482229141
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.