LUCI - N & P Export Coefficents: Tuapaka Catchment

This case study provides an example of four applications of the Land Utilisation and Capability Indicator (LUCI) in water quality modelling to the Tuapaka catchment, a largely agricultural area located to the east of Palmerston North.

Each application uses increasingly detailed input data and the results are compared to OVERSEERTM predictions, as well as instream nitrogen (N) and phosphorus (P) measurements.        

LUCI is applied in this case study to help address the growing need for New Zealand farmers and land managers to reduce nutrient run off leaving their land into surrounding waterways. This change in land use is in response to recent changes to legislation which aim to improve water quality throughout New Zealand.

The LUCI model is well suited to explore options in this setting as it is a GIS framework that considers the impacts of land use on multiple ecosystem services in a holistic and spatially explicit manner. The fact that LUCI was able to model at the resolution of the underlying input DEM meant that its application could both be at a catchment scale as well as providing detailed information for individual farms.            

The four modelling applications to the catchment used increasing detail and catchment specific input data with each subsequent application becoming being more comprehensive than its former;

  • Application 1 used only the default national datasets.
  • Application 2, nationally available spatially varying annual average rainfall and evapotranspiration data by NIWA (Tait et al. 2006; Woods et al. 2006) was replaced by raster surfaces derived from actual rainfall and evapotranspiration data collected from June 2013-June 2014. Derivation was achieved by applying the difference between actual and modelled climate variables at the point of measurement to the NIWA raster climate surfaces.
  • Application 3 used the climate surfaces from 2 with the addition of the Massey University soil map for the Tuapaka Agricultural Experimental farm. This provided more spatial detail around soil variability.
  • Application 4 used the climate surfaces based on actual data, the detailed soil data and actual farm input information from the OVERSEER® xml files. In this application outputs from the LUCI water quality models, including maps and in-stream loads, were then compared to actual water quality data and OVERSEER® predictions.



LUCI water quality modelling provided maps and data for both water quality instream as well on land nutrient loads (N&P). Outputs are presented as a table and series of maps which provided comparisons between each scenario of increasing detail as well as to the OVERSEERtm model and measured results. Map outputs included; nitrogen loads, accumulated nitrogen, phosphorus loads & accumulated phosphorus.

As can be seen in the map outputs, the quality of input data has a very noticeable effect on the quality and resolution of the outputs that LUCI can produce. The addition of actual farm nutrient inputs as well as soil classifications, as in applications 3 & 4, produced more detailed representations of nutrient levels and pathways. These results highlighted where opportunities exist to implement mitigation strategies to intercept nutrients before they enter local waterways.

LUCI accumulated nitrogen output

Accumulated nitrogen load maps from LUCI Application 1 (3a), Application 2 (3b), Application 3 (3c) and Application 4 (3d)

LUCI phosphorus load output

Phosphorus load maps from LUCI Application 1 (4a), Application 2 (4b), Application 3 (4c) and Application 4 (4d)


Modelling showed that the highest P loads were sourced from steeper pastoral grassland while the lowest loads in the catchment came from forested land and flatter land in the upper sections of the catchment. Modelled results were significantly different to measured data indicating room for refinement in the modelling process.

LUCI modelled results were however very similar to prior OVERSEERtm results considering applications 1&2. Notable differences between the two were modelled in applications 3&4 suggesting that additional detail used in these LUCI scenarios may refine the modelling significantly.

LUCI table compares results

 Measured and modelled specific load for the Tuapaka catchment. Note for Row 1 mean is presented with range in brackets.

This case study using LUCI in the Tuapaka catchment highlighted the effect of data resolution and detail on N and P exports. This application has shown that using data specific to a catchment or farm is preferable when using LUCI if the data is available.

While LUCI was comparable to existing models in terms of outputs, clear differences still exist between measured nutrient losses and LUCI predictions at the catchment scale. Improved understanding of nutrient attenuation in the catchment is likely to improve the accuracy of LUCI predictions.

Further modelling improvements as shown in this case study include the addition of detailed soil data (Application 3). This had a clear impact on sources and loads of P within the catchment. Detailed farm input data also decreased loads for both N and P. These differences indicate that using data specific to a catchment or farm is preferable for use in LUCI. Additionally, this highlights the ease with which actual and specific data can be incorporated for use in LUCI.     

Associated Models

Land Utilisation and Capability Indicator (LUCI)

The Land Utilisation and Capability Indicator (LUCI) model allows the mapping of ecosystem services from the sub-field to national scale, and modelling of the impact of management services on these services.


Trodahl, M., Burkitt, L., Bretherton, M., Deslippe, J., Jackson, B. & Metherell, A., (2017): Developing N & P Export Coefficients for Rural Landscape Modelling in LUCI. In: Science and policy: nutrient management challenges for the next generation. (Eds L.D. Currie and M.J. Hedley). Occasional Report No. 30. Fertilizer and Lime Research Centre, Massey University, Palmerston North, New Zealand. 9 pages.

Woods, R., Hendrikx, J., Henderson, R., & Tait, A. (2006). Estimating mean flow of New Zealand rivers. Journal of Hydrology (New Zealand), 95-109.

Tait, A., Henderson, R., Turner, R., & Zheng, X. (2006). Thin plate smoothing spline interpolation of daily rainfall for New Zealand using a climatological rainfall surface. International Journal of Climatology, 26(14), 2097-2115.