Reformulation of the Drought Code in the Canadian Fire Weather Index System Implemented in New Zealand

In the Canadian fire weather index (FWI) system implemented in New Zealand, the drought code (DC) is a daily index of water stored in the soil and is calculated using surface air temperature, daily rainfall, and a daylength factor. In this study we propose two methods to improve the DC calculation. The first one (called PotE) employs a potential evaporation scheme that considers wind speed, surface air stability, and water mixing ratio gradient. The second one (called soilM) uses the ratio of the relative soil moisture (with respect to the wilting point at which soil moisture stress completely prevents transpiration) to the difference between the soil moisture at saturation and the soil moisture at wilting point. One advantage of soilM over PotE and the old scheme is that the effect of different soil types is considered in the soilM scheme. Another advantage of soilM is that when the soil moisture is from observation, the calculated DC represents the actual drought status. We have calculated the DC and FWI with the original DC scheme, PotE, and soilM using observed soil moisture (called obs_soilM) and simulated soil moisture (called sim_soilM) from the joint UK land environment simulator (JULES) at 28 climate stations in New Zealand from 1 Feb. 2009 to 31 Jan. 2011. The first year was used to spin up the system and the second year for analysis. The results showed that the spinup of the DC may take up to half a year. The original DC scheme severely underestimated the drought status in New Zealand, especially in summer. This led to underestimation of the FWI in summer. The errors in the calculated drought status and FWI were significantly reduced by the PotE and sim_soilM schemes, especially the latter. Although soil moisture observations are rare, simulated soil moisture can be easily obtained by running a land surface model using surface observations or analysis. This indicates the potential application value of the suggested sim_soilM scheme in drought status and FWI calculation and predictions. For all the three DC schemes, the calculated DC and FWI are more sensitive to rainfall, and less sensitive to air temperature at most of the 28 stations. The sensitivity to rainfall mainly occurred in warm seasons. The magnitude of the sensitivity varies among stations due to differences in local climate. The sensitivity of each scheme used to calculate the DC is almost the same.