Global dynamics is influenced by many factors, which can make it very irregular. The development of the epidemic in some countries may be more similar to
the picture described by SIR model. But the corresponding picture can be obtained
only if the registered number of cases is close to the real one. In countries where the
problems with testing (and, accordingly, with the isolation of patients) occurred,
one should not expect good simulation results. Unfortunately, Ukraine is one of
such countries, so even the eighth forecast (see Table 6.23) made using data for the
period T c immediately preceding the first explicit maximum of the second derivative on May 16 (visible in Fig. 8.2) did not provide good accuracy (e = 89%).
We can threat the days: May 16, May 29, June 8, July 3 and 19, August 2 and
18; the end of August and September as the beginning of the second, third, etc.,
waves of the COVID-19 pandemic in Ukraine (see Fig. 8.2). Some of these days
are located very close, which reduces the number of observations of the corresponding waves and the accuracy of calculations. So in papers [89, 94], some of
these waves were combined. The presence of epidemic waves in Ukraine is also
evidenced by two clear minimums of dV/dt (see “triangles in Fig. 8.2), but the
corresponding values are very high and do not differ very much from the maximum
daily number of new cases typical for early May and mid-June. We will focus on
the analysis of the next epidemic waves in Ukraine in Chap. 11.
The proposed method is very sensitive to data irregularities and can be also used
for their detection. For example, the number of accumulated cases in the UK
(reported to WHO on May 20) was lower than on May 19. The same situation
occurred between July 1 and 2. The corresponding irregularities in the values of the
first and second derivatives are visible in Fig. 8.3. Neglecting them, we can
Fig. 8.13 Pandemic dynamics in the Republic of Moldova versus time in days. Accumulated
number of cases (V j and w j —“circles”; smoothed values—line, Eq. (8.1)). “Triangles” show the
first derivative (Eqs. (8.2) and (8.4)) multiplied by 100, “stars”—the second derivative (Eqs. (8.3)
and (8.5)) multiplied by 1000
118
8 Identification of the New Waves of the COVID-19 Pandemic
the picture described by SIR model. But the corresponding picture can be obtained
only if the registered number of cases is close to the real one. In countries where the
problems with testing (and, accordingly, with the isolation of patients) occurred,
one should not expect good simulation results. Unfortunately, Ukraine is one of
such countries, so even the eighth forecast (see Table 6.23) made using data for the
period T c immediately preceding the first explicit maximum of the second derivative on May 16 (visible in Fig. 8.2) did not provide good accuracy (e = 89%).
We can threat the days: May 16, May 29, June 8, July 3 and 19, August 2 and
18; the end of August and September as the beginning of the second, third, etc.,
waves of the COVID-19 pandemic in Ukraine (see Fig. 8.2). Some of these days
are located very close, which reduces the number of observations of the corresponding waves and the accuracy of calculations. So in papers [89, 94], some of
these waves were combined. The presence of epidemic waves in Ukraine is also
evidenced by two clear minimums of dV/dt (see “triangles in Fig. 8.2), but the
corresponding values are very high and do not differ very much from the maximum
daily number of new cases typical for early May and mid-June. We will focus on
the analysis of the next epidemic waves in Ukraine in Chap. 11.
The proposed method is very sensitive to data irregularities and can be also used
for their detection. For example, the number of accumulated cases in the UK
(reported to WHO on May 20) was lower than on May 19. The same situation
occurred between July 1 and 2. The corresponding irregularities in the values of the
first and second derivatives are visible in Fig. 8.3. Neglecting them, we can
Fig. 8.13 Pandemic dynamics in the Republic of Moldova versus time in days. Accumulated
number of cases (V j and w j —“circles”; smoothed values—line, Eq. (8.1)). “Triangles” show the
first derivative (Eqs. (8.2) and (8.4)) multiplied by 100, “stars”—the second derivative (Eqs. (8.3)
and (8.5)) multiplied by 1000
118
8 Identification of the New Waves of the COVID-19 Pandemic