Sample handling procedures at all NADP/NTN sites were changed substantially on 11 January 1994 in order to reduce contamination from the sample shipping container. This notification alerts you that the data for samples before and after that date are different and not comparable. See the table below for a tabular summary of the differences, based on a special intercomparison study. The text below describes the rationale for the change of procedures.

Prior to 11 January 1994, field samples were sent to the NADP/NTN Central Analytical Laboratory (CAL) in the high density polyethylene buckets used for collection. Only a 20-mL aliquot, used for field chemistry measurements, was removed before the sample was sealed and sent to the CAL. This procedure had the advantage of minimizing the handling of samples at field sites and of allowing the CAL to measure and record what was in a virtually complete sample. Site operators’ sample volume measurements and reports of contaminants could be verified at the CAL and amended as needed. Despite these advantages, a small but important source of contamination was found in the rubber o-ring that sealed the lid to the sample bucket, preventing leaks during shipment. This o-ring was a source of the same cations and anions that are measured in wet deposition. Most affected were free hydrogen ion concentrations at sites with low ionic strength precipitation, causing field pHs to be several tenths of a pH unit lower than CAL pHs at many western U.S. sites.

Efforts to improve lid cleaning procedures reduced but could not eliminate the contamination. This persistent contamination and occasional changes in the o-rings resulted in systematic time- varying biases that have the potential to interfere with the interpretation of NADP/NTN data and long-term trends studies. Special studies had shown that samples sent in bottles have a much smaller loss of hydrogen ions than samples sent in the collection buckets. As a result, the NADP/NTN elected to change the field protocol and use bottles to send samples to the CAL, pending the outcome of an intercomparison study of old and new methods at 11 pre-selected sites.

This intercomparison study began in 1992 and ended in December 1993. Samples were collected in side-by-side collectors set to open and close together. One sample was handled the old way and sent to the CAL in the collection bucket. The other sample was handled according to the new protocol and was taken from the field site to the field laboratory in a bucket with a snap-on lid and no o-ring. At the field laboratory the sample was poured into a one- liter, wide-mouthed, high density polyethylene bottle and sent to the CAL. Sample in excess of one liter was discarded and an effort was made not to transfer contaminants (floating, suspended, or settled) to the bottle. Results from this study confirmed the larger loss of hydrogen ions in samples sent the old way to the CAL, and the NADP/NTN implemented the protocol change on 11 January 1994.

Results of the intercomparison study are listed in Table 1.

If you have questions or comments please contact the

NADP Program Office,
Wisconsin State Laboratory of Hygiene
465 Henry Mall
Madison, WI 53706

Table 1.  Results of Intercomparison of Old and New Sample Handling Procedures at NADP/NTN Sites.

Median and 95% Confidence Interval of Change (1)

Number of
Sample Pairs
Direction of
Concentration (mg/L) Concentration
Sulfate405decrease0.02  +/- 0.0042.6  +/- 0.5
Nitrate405decrease0.01  +/- 0.0040.9  +/- 0.4
Ammonium405no change0.00  +/- 0.0030.0  +/- 1.2
Calcium405decrease0.002 +/- 0.0012.5  +/- 1.2
Sodium405decrease0.005 +/- 0.0029.9  +/- 2.5
Chloride405no change0.00  +/- 0.0020.0  +/- 1.1
Magnesium405decrease0.002 +/- 0.00039.5  +/- 1.6
Potassium405decrease0.002 +/- 0.000511.1 +/- 3.1
H+ for pH <4.6112increase3.2 +/- 1.17.2 +/- 2.3
H+ for 4.6<=pH<=5.6191increase2.5 +/- 0.535 +/- 7
H+ for pH> 5.6102increase2.1   +/- 0.8220  +/- 91
(pH Units)(Percent)
H+ for pH <4.6112increase0.03 +/- 0.01n/a
H+ for
191increase0.13 +/- 0.02 n/an/a
H+ for pH> 5.6102increase0.50  +/- 0.10n/a
Conductance405increase0.9  +/- 0.19.3  +/- 1.7

(1) Changes are the median differences between the old (before 11 January 1994) and new protocols. Where a direction of change is listed, the null hypothesis that the two protocols resulted in the same concentrations could be rejected at the 0.05 level (Wilcoxon signed rank test). The 95% confidence interval was calculated from the median [1.58 (C75% – C25%)/SQRT(n)] (see Velleman and Hoaglin, 1981: Applications, Basis, and Computing of Exploratory Data Analysis, Duxbury Press, Boston, MA 02116).