Skip to main content

Effects of land use change on the flood hydrology of the Waikato River catchment

Introduction

Increasing pressure is being exerted for conversion of established areas of forest plantation to pasture in the Waikato River Catchment. Such conversion is already well underway in some parts of the catchment. 

Waikato Regional Council has been aware of the gradual conversion within the upper Waikato catchment over the last several years. Due to the scale of land use change, a policy review for the area above Karapiro is proposed. The review process requires specific steps be followed including scientific investigations of the issues and their effects on current policies and targets, available methods to address the issues and effects including justifications for any policy changes. Stakeholder and public consultation is a legal requirement and is part of the whole review process. The initial assessment of the effects of land use change within the catchment above Karapiro included effects on water quality, soil erosion and flood hydrology. This project is concerned only with understanding the effects of forest plantation to pasture conversion on flood hydrology within the Waikato River Catchment downstream of Lake Taupo. 

These areas consist of a variety of geological soils, whose hydrological and erosion characteristics vary with changes in land use. In particular the flood hydrology of the pumice soils of the central North Island is known to be particularly sensitive to changes in land use. Conversion of plantation forested areas to pasture within the Waikato region is currently unregulated by statutory planning instruments and can occur ‘as of right’. Currently known areas of forestry conversion are assessed at up to approximately 70,000ha.

Project outline

You can download the project brief and plan below to learn about the effects of land use change on the flood hydrology of the Waikato River catchment between Karapiro and Taupo. 

Report

The Waikato River at Rangiriri drains a catchment of approximately 12420km2.  This study looks at changes to the upper catchment between Karapiro and Taupo which comprises an area of 4400km2. Within this upper catchment, 52% of land cover is exotic forest, indigenous vegetation, scrub, or unmanaged areas, while   45.7% is being used for agricultural purposes.  The potential conversion of 567 km2 of forest (24% of the existing forested land) to pastoral agriculture over the next 15 years represents an area of land use change of 12% of the total land area of the Taupo to Karapiro catchment; this is equivalent to 4% of the catchment that drains to Rangiriri. 

Any significant change in land use has potential to impact on flood hydrology of the Waikato River and its tributaries. This includes subsequent effects downstream of Karapiro dam and impacts on the flood protection works of the Lower Waikato Waipa Control Scheme. 

Previous scientific studies within the Waikato catchment on the effects on flood size of conversion from forest to pastoral agriculture indicate that both the rate and total volume of flood runoff increase. The studies were in very small catchments and cannot be extrapolated to large catchments. The magnitude of these observed increases in flood peaks ranged from a factor of two to ten. The explanation given for this change is the reduction in the infiltration capacities of the soil following conversion to pastoral agriculture as a result of soil compaction due, for example, to grazing animals and vehicle use. Thus, while scientific consensus exists for an increase in flood flows from pasture, the increases are highly variable and based on limited and small-scale studies. This necessitated a concerted modelling study at the regional scale as presented herein. 

The Waiakto Regional Council established a project in 2007 to assess the effect that the anticipated potential changes in land use may have on the flood hydrology of the Waikato River and its tributaries. The scope of the project was defined by a Project Brief, and overseen by both a Project Control Group and a Technical Expert Panel. 

This report assesses the change in Waikato flood hydrology for different magnitude floods at three different scales: (i) local flooding within the upper Waikato; (ii) Karapiro outflows, and (iii) the lower Waikato River system. The floods investigated here range in size from small to extreme, and correspond to rainfall events of specified return intervals. This range of event sizes enables the study to produce assessments that are relevant to a wide range of possible impacts, ranging from small relatively frequent floods to design and over-design events. These will assist stakeholders in their respective responsibilities for long-term planning. However, it must be stressed that the specific return intervals for the rainfall events do not translate to the floods.

The major steps in the study were to summarise data and research on floods and land use in the region, and to use simulation models to predict the impacts of land use change in the Upper Waikato on flood hydrology.

The simulation modelling comprised three steps:  

  1. Modelling of the flood response of the seven hydrolake sub-catchments, using two separate models;
  2. The propagation of the resulting flood responses through the hydrolake system; and
  3. Assessment of how these flood pulses affect inundation in the Lower Waikato.

The first step provided the basis for the study. It comprised the use of two distinct catchment hydrology models to predict the flood response to storms of different magnitude. These predictions were made twice, for each model: once for the current land cover, and once for potential future cover following forest-to-pasture conversion. The differences between the pre- and post-conversion scenarios thus provided an estimate of how the forest conversion might affect flooding.

The modelling approach was employed because observational data alone were insufficient to identify the effects of land cover differences. The majority of the historical forest-pasture conversion pre-dates even the longest of flow records, and the flow observations are mostly from catchments with relatively similar land cover composition. 

Data used to underpin the modelling study comprised: 

  • catchment land use, derived from the Land Cover Database (LCDB2),
  • geology,
  • soil characteristics, derived from the New Zealand Land Recourse Inventory (NZLRI),
  • hydrological data, including streamflow and rainfall records as well as in situ observational studies,
  • hydraulic data for the Lower Waikato, and
  • dam operational information.

The authors consider that this report presents a defensible approach to the estimation of land-use change impacts on floods at the catchment scale of interest.  Factors contributing to our level of confidence include:

  1. a reasonable match between modelled and measured flood response under current land use
  2. results are consistent with physically-based reasoning (e.g., larger impacts on small catchments with large fraction of the land cover changing)
  3. results for changes in local flooding are consistent with measurements in small catchments
  4. limitations in the data available to validate change predictions for large catchments

 

The modelling study found that:

  1. Simulation models provide a useful means to investigate how land use affects the transformation of rainfall into flood runoff; and
  2. The projected changes in land use would lead to the following changes in flood peaks:

 

 

Modelling study table

 

Small flood
(5-year rainstorm)

Medium flood
(20-year rainstorm)

Large flood
(100-year rainstorm)

Extreme flood
(500-year rainstorm)

Local flooding  within Upper Waikato

10-100 km2 catchment area,
0-80% upstream land use conversion

Significant increase (5-50%) for streams where most of catchment has land use change

Significant increase (5-50%) for streams where most of catchment has land use change

Very significant increase (more than 50%) for streams where most of catchment has land use change

Very significant increase (more than doubled) for streams where most of catchment has land use change

Upper Waikato Taupo-Karapiro inflow
4405 km2 area,
542 km2 land use conversion (12%)

Little or no change

Little or no change

From 2-9% increase in peak flow rate (average 4%)
0-5% increase in 72-h flood volume (average 2%)

From 2-16% increase in peak flow rate (average 6%)
2-10% increase in 72-h flood volume (average 4%)

Upper Waikato Karapiro outflow

7852 km2 area

Little or no change

Little or no change

From 0.5-3% increase in peak flow rate (mean of 2%)

From 1-12% increase in peak flow rate (mean of 7%)

Waikato River at Hamilton 

8230 km2 area

Little or no change

Little or no change

0-110 mm water level increase
0-21 m3s-1 peak flow increase

0-530 mm water level increase
0-140 m3s-1 peak flow increase  

Waikato River at Ngaruawahia

11395 km2 area

Little or no change

Little or no change

0-40 mm water level increase
0-18 m3s-1 peak flow increase

0-270 mm water level increase
0-150 m3s-1 peak flow increase

Waikato River at Huntly

12066 km2 area

Little or no change

Little or no change

0-40 mm water level increase
0-17 m3s-1 peak flow increase

0-220 mm water level increase
0-150 m3s-1 peak flow increase

Waikato River at Rangiriri

12420 km2 area

Little or no change

Little or no change

0-30 mm water level increase

0-17 m3s-1 peak flow increase

Flood exceeds design standards even under current land use; stopbanks overtopped

 
The study found that the projected change in land use is expected to cause an increase in flood risk during large and extreme rainstorms (that is, 100-year and 500-year rainstorms) that contain intense rain bursts.

The study also found that, consistent with previous small-catchment studies, local flooding may increase significantly for streams where most of the catchment’s land use is converted, but the exact magnitude of change will depend on site-specific details that were beyond the scope of this report. Increases in hydrolake inflows are likely to be greater where fed by sub-catchments with greater portions of forest under conversion, and with smaller coverage of pumice soils beneath the converted forest.

The uncertainty ranges presented in the summary table above reflect the range of model results and rainfall inputs chosen. The uncertainty ranges presented in the report thus reflect a combination of uncertainty in our knowledge of hydrological behaviour and of assumed natural variability. The methods we have adopted provide the best available bounds for what could happen.

The overall conclusions from the study as agreed by the Technical Expert panel are that the effect on flood flows and water levels from land use change in approximately 12% of the Upper Waikato catchment are likely to have:

  • Significant to very significant increases in peak flow rate for local flooding in small catchments where full conversion is expected.
  • While the 72 hour storm simulated is appropriate for defining the effects of land use change over the whole catchment, the effects on the local tributary sub-catchments are larger for storms with shorter durations. To make assessments of impacts on local flooding in specific cases, a range of design storms with different durations need to be considered, to assist in identifying the magnitude of the local effects and appropriate mitigation measures
  • At Hamilton, insignificant impacts during small to medium floods, increases of up to 40-110 mm in peak water level for large floods, and increases of 280-530 mm for extreme floods
  • From Ngaruawahia to Rangiriri, insignificant impacts during small to medium floods, increases in the peak flood water level of 20-40 mm during large floods, and increases of 170-270 mm in extreme floods;

The assumption that land use change will take place on 567 km2 is a realistic scenario, rather than a precise prediction of what future land use changes will occur. When further information becomes available on actual or planned forest conversion in the catchment, the methods developed here can readily be applied with the new land use information. Other potential changes in land use were outside the scope of this report.

Mitigation measures such as flood detention dams can reduce increases in flood size associated with conversion from forest to intensive agriculture, and are being used in the Waikato catchment. The methods used in this study could be extended to include effects of mitigation, if required, but this was outside the scope of the study.

Assessing the significance of these level and flow changes in terms of design and policy considerations was not within the scope of this report.