Costs of measuring SOC

The cost of measuring SOC largely depends on the number of samples, field accessibility, and laboratory analysis. The number of samples to be taken in a project depends on the level of variability in soil organic carbon in the target area, the required levels of precision and resource availability. Soil often has greater spatial variability that demands more sampling efforts. In some cases the cost of demonstrating the change in carbon stocks in soils to the required accuracy and precision may exceed the benefits that accrue from the increase in stocks (IPCC, 2003; MacDicken, 1997). Therefore developing locally calibrated models that can use easily collected data can minimize the cost of demonstrating a change in soil organic carbon stock (IPCC, 2003). Thus, developing alternative cheaper and repeatable measures is a research priority.

Infrared spectroscopy offers promise for a rapid, reliable and cost effective measurement of soil organic carbon. In this study we compared the cost of measuring SOC analyses using the conventional Thermal Scientific FlashEA 1112 CN analyzer, a commercial laboratory in the UK, and near-infrared spectroscopy. All costs are based on conducting sampling and analysis in Kenya.

The total costs of measuring cabon using the Thermal Scientific FlashEA 1112 CN analyzer is USD 20.77 per soil sample, of which 87% are personnel costs (Figure 1b). If acidification is applied to remove carbonates from the sample, the price will increase to USD 25.76. Soil sampling constitutes the highest proportion of the costs of carbon measurement.

Figure 1. Cost of measuring SOC stocks: (a) costs and (b) cost structure of measuring one soil sample. The personnel costs are more or less uniform for the four major activities, while soil sampling constitutes a large proportion of the other costs.

The laboratory costs (without field sampling and sample preparation costs) of measuring carbon using the conventional thermal oxidation and NIR soil spectroscopy are USD 4.99 and 2.19 per sample, respectively. Compared to the thermal oxidation method, soil spectroscopy can reduce laboratory costs of measuring carbon by 56% (Figure 2a). However, there is no significant difference in the total cost of measuring SOC between the Thermal Scientific FlashEA 1112 CN analyzer and the NIR soil spectroscopy when a small number of soil samples are used (Figure 2b). This is because a large proportion of the costs of soil carbon measurement are incurred for soil sampling and preparation compared with laboratory costs (Figure 2a). With increasing number of soil samples, however, the total cost of carbon measurement using NIR spectroscopy is cheaper than using the Thermal Scientific FlashEA 1112 CN analyzer (Figure 2b). Compared to costs of other commercial soil labs, the cost of measuring SOC using the NIR spectroscopy is significantly cheaper than commercial soil laboratories charges (Figure 7.2c, d). However the big advantage of IR technology is the high throughput achievable, which is critical for carbon inventories at project level or larger geographical extents. The daily throughputs of a thermal analyser is quite low (30 samples acidified, or 60 samples unacidifed) whereas NIR throughput is 350 samples per day, and over 1000 per day with robotic MIR systems (Shepherd & Walsh, 2007). Thus throughput rate is the critical determining factor.

Figure 2. Comparisons of costs of measuring SOC carbon: (a) laboratory and (b) total costs of measuring SOC using the Thermal Scientific FlashEA 1112 CN analyzer and NIR soil spectroscopy. Both the (c) laboratory and (d) total costs of measuring SOC using NIR spectroscopy is significantly cheaper than the costs of measuring it in commercial soil labs.

Cost-error analysis

According to the Marrakesh Accords, uncertainties in measuring greenhouse gases in offsetting projects should be quantified. Estimation errors, model errors, and sampling errors associated with the number of samples are among the major sources of uncertainties in measuring SOC. IPCC (2003) has recommended using confidence intervals as quantitative estimate of uncertainty.

To estimate the sample size required to measure carbon stocks with the desired confidence interval (95%), we used the mean (21.53 t C ha-1) and standard deviation (13.62 t C ha-1) of carbon stocks in the top soil (20 cm) of the five sites in the western Kenya (Figure 3).

To decrease the confidence interval from 4.03 to 2.02 t C ha-1, the sample size should increase from 50 to 200 (Figure 3a). The increase in the number of samples from 50 to 150 in turn increases the cost of carbon measurement by USD 3115 (Figure 3b).

Figure 3. (a) 95% confidence intervals (t C ha-1) of the carbon stock and (b) the measuring cost of carbon in the topsoil layer (0-20 cm) from 160 plots in western Kenya.

__________

IPCC. 2003. Good Practice Guidance for Land Use, Land-Use Change and Forestry, In Penman, J., et al., eds. Institute for Global Environmental Strategies (IGES), Japan.

MacDicken, K.G. 1997. A Guide to Monitoring Carbon Storage in Forestry and Agroforestry Projects. Winrock International, Arlington, Virginia.

Shepherd, K.D., & Walsh, M.G. 2007. Infrared spectroscopy-enabling an evidence-based diagnostic surveillance approach to agricultural and environmental management in developing countries. Journal of Near Infrared Spectroscopy 15:1-19.